>

PROCEEDINGS

OF THE

AMERICAN ACADEMY OF ARTS AND SCIENCES.

PROCEEDINGS

OF THE

AMERICAN ACADEMY

OF

ARTS AND SCIENCES.

NEW SERIES. Vol. XIV.

WHOLE SERIES. Vol. XXII.

FROM MAY, 1886, TO DECEMBER, 1886.

SELECTED FROM THE RECORDS.

BOSTON:

UNIVERSITY PRESS: JOHN WILSON AND SON.

1887.

Xb'lo I

CONTENTS.

Page I. Observations of the Mean Right Ascension of Certain Polar

Stars, made at the Field Memorial Observatory of Williams

College, and reduced to the Epoch 1SS4-0. By Truman

Henry Safford 1

II. Thunder- Storms in New England in the Summer of 1885. By

William Morris Davis 14

III. A Preliminary Synopsis of North American Carices, including

those of Mexico, Central America, and Greenland, with the American Bibliography of the Genus. By L. H. Bailey, Jr. 59

IV. Contributions from the Chemical Laboratory of Harvard College.

I. Analysis of Mica from Leon Co., Texas. By George W. Leighton 158

II. Analysis of a Crystalline Scale formed in the Manufacture of Sodic Bicarbonate by the Ammonia Process at Syracuse,

N. Y. By George W. Leighton 159

III. On the Constancy in the Heat produced by the Reaction of Argentic Nitrate on Solutions of Metallic Chlorides. By Theodore W. Richards 162

IV. Is there a Constant Relation between the Heats of Forma- tion of Chlorides and Sulphates in A queous Solution ? By Irving W. Fay 164

V. A Method for the Separation and Estimation of Boric Acid, with an Account of a Convenient Form of Apparatus for Quantitative Distillations. By F. A. GoocH 167

vi CONTENTS.

Page VI. A Method for the Separation of Sodium and Potassium from

Lithium by the Action of Amyl Alcohol on the Chlorides,

with some Reference to a similar Separation of the same from

Magnesium and Calcium. By F. A. GooCH 177

Vn. On the Anatomy and Development of Agarum Turneri. By

James Ellis Humphrey 195

VIII. On the Morphology of Kavenelia Glandulaformis. By G. H.

Parker 205

IX. Experiments on the Melting Platinum Standard of Light. By

Charles R. Cross 220

X. The Inverse Electromotive Force of the Voltaic Arc. By

Charles R. Cross and Wm. E. Shepard 227

XI. On the Direct Conversion of Aromatic Sulphonates into the Cor- responding Ainido Compounds. By C. Loring Jackson AND John F. Wing 245

XIT. On Benzyldimethylamine. By C. LoRiNG Jackson and John

F. Wing 249

XIII. A New Method for the Quantitative Determination of Hydroxyl.

By C. Loring Jackson and G. W. Rolfe 253

XIV. On Certain Cultures of Gymnosporangium, with Notes on their

JRoestelice. By Roland Thaxter 259

XV. Contributions to American Botany. By Asa Gray . . . 270

XVI. On Mucoxybromic and Mucoxychloric Acids. By Henry B.

Hill and Arthur W. Palmer 315

XVII. A Determination of the Relation of the Atomic Weights of

Copper and Silver. By Theodore W. Richards . . 342

XVIII. On Benzoltrisulphonic Acid. By C. Loring Jackson and

John F. Wing 348

XIX. On the Action of Nitric Acid on Symmetrical Trichlorbenzol.

By C. Loring Jackson and John F. Wing . . . 372

XX. Observations of Variable Stars in 1886. By Edward C.

Pickering 380

CONTENTS. VU

Page

XXI. Contributions to American Botany. By Sereno Watson . 396

XXII. On the Behavior of Sound and Decayed Wood at High Temperatures. By Henry B. Hill and Arthur M.

COMEY 482

XXIII. The Efficiency of Small Electro-motors. By H. E. H.

Clifford 493

Proceedings 499

Memoirs: —

Charles Francis Adams 517

Nathaniel Ellis Atwood 522

Ephraim Whitman Gurney 523

William Ripley Nichols 528

Charles Callahan Perkins 534

Henry Hobson Richardson 539

Leopold von Ranks 542

List of the Fellows and Foreign Honorary Members . . 559

Index 567

A

OtM.

T-

Rec Ac< Civ Pla

■•;-H:-i^•^^*^;.-rif■iaJP???s?sf^^

PROCEEDINGS

OF THE

AMERICAN ACADEMY

OF

ARTS AND SCIENCES.

VOL. XXII. PAPERS READ BEFORE THE ACADEMY.

OBSERVATIONS OF THE MEAN RIGHT ASCENSION OF CERTAIN POLAR STARS, MADE AT THE FIELD MEMORIAL OBSERVATORY OF WILLIAMS COLLEGE, AND REDUCED TO THE EPOCH 1884.0.

By Truman Henry Safford.

Presented April 14, 1886.

The observations here presented were made mostly in 1884 ; earlier were not before reduced, for reasons of convenience ; verv few later, in 1885, are here added for completeness' sake.

MARINE BIOLOGICAL LABORATORY.

^ • -^

ReceWed ^J^^i.^^ /o '^ / ry o Accession No. ^ .■^\'~~~ Given by^^T^.C?^.'^?rr?^r^c^^^ /^^ Place,

***Ho book OP pamphlet is to be Pe moved from the Iiab- opatopy luithout the pepmission of the Tpustees.

a few and a

and 1 be 'ight con- ;ame

this ;iven dons

t the ,s of My n of tion, :e of

PROCEEDINGS

OF THE

AMERICAN ACADEMY

OF

ARTS AND SCIENCES.

VOL. XXII. PAPEHS READ BEFORE THE ACADEMY.

OBSERVATIONS OF THE MEAN RIGHT ASCENSION OF CERTAIN POLAR STARS, MADE AT THE FIELD MEMORIAL OBSERVATORY OF WILLIAMS COLLEGE, AND REDUCED TO THE EPOCH 1884.0.

By Truman Henry Safford.

Presented April 14, 1886.

The observations here presented were made mostly in 1884; a few earlier were not before reduced, for reasons of convenience ; and a very few later, in 1885, are here added for completeness' sake.

The latter portion of the observations of 1884, not here given, and those of 1885, are in progress of reduction to 1885.0, and will be offered to the Academy later. The present series includes right ascensions only, as with the instrument here employed it is not con- venient to observe both co-ordinates of so close polars at the same time. The region in which I am at present working allows this simultaneous observation. The declinations of the stars here given are readily observed along with both right ascensions and declinations of the stars now taken.

The series of which this work is a part was begun in 1882, at the Field Memorial Observatory of Williams College. The results of 1882-83 are given in the Proceedings, Vol. XIX. pp. 324-352. My intention was and is to carry it over a pretty extensive region of the heavens, taking in from time to time stars needing observation, and those only. This condition, it is true, offers a wide range of

VOL. XXII. (N. S. XIV.) 1

2 PROCEEDINGS OF THE AMERICAN ACADEMY

choice ; but the attempt is made to avoid needless duplication of the work of others. A previous series of right ascensions, which I ob- served in 1862-G6 at Cambridge, was not found superfluous.

Both then and now I have been impressed with certain deficiencies in our fundamental catalogues, as, indeed, have other astronomers. These deficiencies are not the fault of those who have constructed the catalogues, but arise from the lack of foresight in the astronomers of the last century. Our fundamental and secondary places are perfectly good and sure, taken as a whole, for 1865 or thereabouts ; but in many cases accurate proper motions are lacking to bring thera up to the present epoch. Stars which Bradley observed with suffi- cient completeness about 1755 can be accurately brought up by proper motions derived from his observations.

But Bradley could not foresee what observers would need a century or more after his decease ; and omitted to provide us all the materials we now need, especially in this polar region. The next set of ob- servers — Piazzi, Groombridge, etc. — did not emploj^ stellar observa- tions for the correction of their instruments, but meridian marks ; and depended upon occasional adjustments, overlooking the tendency now so well known of instruments to crawl away from their adjustments through molecular strains connected with change of temperature. Consequently, when Bradley, 1755, fails us, the observations of Struve, beginning in 1814, are the first which really come up to Bradley's standard.

So that in my use of the best fundamental catalogue now existing, that of the Astronomische Gesellschaft, I have thought it safest to begin by excluding for instrumental corrections all stars whose proper motions do not essentially depend upon Bradley, or upon Struve and later observers. This excludes quite a number of the right ascensions even of the Hauptsterne ; as well those whose errors Professor Rogers and Mr. Chandler have detected, as those which, equally uncertain a pi'iori, are approximately con-ect by compensation of errors. In their plac^ I have substituted a number of closer polars whose right ascensions are thoroughly well determined by modern observations, and whose proper motions are also accurately known. In collecting modern observations, my own Harvard catalogue of 1865, and the work since done there by Professor Rogers, as well as the Williams- town observations, have been added to the ordinary catalogues ; and I have sought out and reduced the single Pulcova observations made in this region by Winnecke and Gromadski between 1858 and 1869.

OP ARTS AND SCIENCES. 3

The right ascensions so obtained are amply sufficient both in quan- tity and quality to replace the excluded " Hauptsterne," although not quite equal to the latter for their own proper epoch, 18G5. The greater nearness of these stars to the pole in certain right-ascensions renders them better for instrumental errors than, for example, such as 7 Urste Minoris, whose right ascension is both uncertain and erro- neous at the present time ; and in fact better than ( Urste Minoris, which is some degrees nearer, equally uncertain, but actually not quite so much in error, so far as I can judge.

It has not been yet necessary to employ any stars whose right ascensions are not given in one or more ephemerides ; so that the following corrections to the apparent places of these ephemerides are currently employed to bring the instrumental corrections into harmony with the Pulcova system, as modified by Auwers.

Two of Auwers's right ascensions are also corrected ; Draconis 1 H., as stated in my previous paper; and 76 Draconis, in conse- quence of an error of sign in the proper motion, which has been since corrected in the Jahrbuch, ,

Groombridge 944	American Ephemeris	—0.38	—0.011	< — 1885)
Camelopardali 25 H.	American Ephemeris	—0.58	—0.008	t — 1885)
Groombridge 1119	Connaissance des Temps	+6.1	+0.32	i — 1885)
Draconis 1 H.	Berliner Jahrbuch	+0.26	+0.0097 <	t — 1885)
Bradley 1672	Conn, des Temps	+1.6	+0.063	<- 1885)
Camelopardali 32 H.	American Ephemeris	—0.30	—0.009	« — 1885)
Groombridge 2283	Conn, des Temps	—2.2	—0.089	t — 1885)
Bradley 2701	Montsouris Ephemeris	+0.2	+0.008	t — 1885)
76 Draconis	B. J. (to 1886 inclusive)	+0.366	+0.0282	[t — 1885)
Groombridge 3548	Conn, des Temps	—0.76	-0.017	t — 1885)
Cephei 32 H.	Montsouris Ephemeris	0.
Cepbei 39 H.	Conn, des Temps	0.

In extending my scheme southward, I have, as will be seen, in- serted the doubtful Hauptsterne, and all the Zusatzsterne, so far as ray zone now goes, into my working list. The doubt refers almost exclusively to right ascension ; hence this co-ordinate will be deter- mined, while at present the catalogue declination is used to find the equator-point.

During the present year I hope to begin a series of experiments with the Rittenhouse* collimator, for the purpose of showing how far

* It seems to be certain that our countryman, David Rittenhouse, was the first to employ an object-glass of small curvature with a meridian mark in its focus. It is quite possible that this application preceded that of an ordinary telescope as a collimator; but the latter is practically a different instrument.

4 PROCEEDINGS OP THE AMERICAN ACADEMY

I can depend upon the stability of the soil in the Observatory grounds. This collimator has an object-glass of five inches in diameter mounted upon a pier south of the circle ; on the same pier are Y's for the other collimator, and the object-glass is between them, but can be turned down so that the telescope of two and one half inches' aperture can be placed in the Y's over it. The telescope, which has pivots and a level, is usually placed upon the north collimator pier, which has Y's, but no object-glass.

In the ordinary position, then, we have two collimators which can be set upon each other; the focal mark is one hundred Paris feet south of the five-inch object-glass, and the north collimator has an excellent micrometer.

"When the north collimator telescope is transferred to the south Y's, it can be used with its level for the determination of the horizontal point, as well as in its usual position.

The whole arrangement is ingenious, original, and economical, and admirably adapted to an observatory whose means are insufficient for strictly primary work, while, with the addition of a few not very ex- pensive auxiliaries, it can be readily adapted to primary work. The object-glass and eye-piece are interchangeable, a requirement which I should always make for any fixed instrument, (I believe the Rep- sokls always construct their large instruments so,) and there is no difficulty whatever in observing the nadir whenever we please.

But at present I have no intention of attempting a strictly primary catalogue ; enough such catalogues are now in course of construction, and I need only to take care that the casual errors of the primary catalogue which I employ do not become systematic errors in my re- sults. Of this there is some danger when the Jahrbuch in its present condition is employed without discriminating between well and ill deter- mined polars, — that is, without reading Professor Auwers's memoir.

The general arrangement of the tables which follow is like that in my previous paper, Vol. XIX. of the Proceedings, pages 324-352. The mean right ascensions are given in two different ways, according as their places are or are not given in one of the great ephemerides. In the former case the datum of each observation is the correction to the ephemeris ; and the mean right ascensions for 1884.0 are obtained by applying to that given in the ephemeris the mean of these correc- tions. For other stars no proper motions are here applied.

As the observations reduced to 1885.0 will soon follow these, I have thought it needless to form a catalogue of mean right ascensions for this paper.

OP ARTS AND SCIENCES.

MEAN RIGHT ASCENSIONS FOR 1884.0. Results of Separate Observations.

Groombridge 67.	Bradley 402.
h. m. 8.		h. m. s.
1881, May 2 s. p. 0 24 50.18	1883, June 22 s. p.	3 5 2.12
3 s.p. 51.04	23 s. p.	2.20
24 s. p. 50.30	1884, June 26 s. p.	2.64
28 s. p. 50.03	July 3 s. p.	3.29
Nov. 21 50.65	7 s. p.	2.82
22 51.38	10 s. p.	2.80
29 50.27	Mean A. R. 1884.0	3 5 2.645
Dec. 2 50.G4
Mean A. R. 1884.0 0 24 50,661	Groombridge 848.
Radcliffe 134.	1884, July 18 s. p. 21 s.p.	B.J. +0.04 +0.00
h. m. 8.	22 s. p.	+0.10
1884, Dec. 2 0 28 32.22	25 s. p.	—0.09
4 32.50		Mean +0.012
29 32.33		B. J. 14.581
30 32.52		h. m. s.
Mean A. R. 1884.0 0 28 32.392	Mean A. R. 1884.0	4 33 14.593
	Radcliffe	1272.
SCHWEED 43.		b. m. s.
h. m. 8.	1884, July 9 s. p.	4 40 28.13
1884, Nov. 15 0 51 13.42	10 s. p.	28.49
22 13.76	11 s.p.	28.23
29 13.49	15 s. p.	28.03
Dec. 2 13.44	Mean A. R. 1884.0	4 40 28.220
Mean A. R. 1884.0 0 51 13.528
	Radcliffe	1311.
Bradley 95.		h. m. 8.
h. m. 8.	1884, July 21 s. p.	4 50 48.17
1882, Oct. 20 0 56 46.11	22 s.p.	48.22
1884, Jan. 10 46.75	25 s.p.	47.34
12 46.58	26 s. p.	47.98
17 47.01	Mean A. R. 1884.0	4 50 47.928
Mean A. R. 1884.0 0 56 46.612
	Groombridge 966. |
Radcliffe 878.		8.
	1884, July 18 s. p.	B.J. +0.04
h. m. 8.	22 s. p.	+0.45*
1884, May 28 s. p. 18 37.05	25 s. p.	+0.04
Nov. 15 36.81	28 s. p.	—0.13
29 37.03	Mean +0.100
Dec. 2 36.91		B. J. 13.185
4 36.27		h. m. s.
Mean A. R. 1884.0 1 8 36.814	Mean A. R. 1884.0	5 24 13.285
* The observation of July 22 seems anomal	ous ; but no error can be found in the wires or |
reductions. The air was thick, and the star p	robably faint and tremulous.	1

PROCEEDINGS OF THE AMERICAN ACADEMY

MEAN RIGHT ASCENSIONS FOR ISSi.O. — Continued.

Gkoombeidge 944.	Camelopardali 25 H.
1884, Feb. 1 A. E. — 0.72	1884, Feb. 15 A. E. —0.37
15 —0.71	Mar. 4 —0.82
July 18 s. p. —0.03	18 —0.96
21 s.p. —0.17	22 —0.78
22 s. p. +0.11	Mean —0.732
25 s. p. —0.40	A. E. 36.827
Aug. 14 s. p. —0.06	h. m. s.
16 s. p. —0.18	Mean A. R. 1884.0 7 6 36.095
18 s. p. —0.18
Mean —0.260
A. E. 56.518
^Ai* .M—^W \^\^»\^^\y	SCHWERD 406.
h. m. s.
Mean A. R. 1884.0 5 24 56.258	h. m. 8.
	1884, Feb. 15 7 9 26.92
	Mar. 4 26.25
Groombridge 956.	18 26.67 22 26 70
Vi m a	25 26.62
1883, Aug. 14 s. p. 6 29 32.75	Mean A. R. 1884.0 7 9 26.632
21 s.p. 33.27
22 s.p. 33.02
1884, July 21 s. p. 83.43	Groombridge 1278,
25 s. p. 33.56
28 s.p. 33.73	h. m. 8.
30 s. p. 33.29	1884, Feb. 15 7 13 28.81
Mean A. R. 1884.0 5 29 33.293	Mar. 4 2&.47
	18 28.61
	22 28.76
	25 28.73
Groombridge 1004.	Mean A. R. 1884.0 7 13 28.676
h. m. 8.
1884, Feb. 1 6 0 55.19 15 55.81	Groombridge 1119.
July 30 s. p. 55.85	8.
Aug. 2 s. p. 55.48 14 s.p. 54.84	1884, Mar. 15 C. des T. +6.46 18 +6.76
15 s. p. 54.68	22 +4.26
Mean A. R. 1884.0 6 0 55.308	25 +4.45
	Mean +5.482
	C. des T. 48.39
Camelopardali 23 H.	h. m. 8. Mean A. R. 1884.0 7 39 53.872
1884, Feb. 1 B. J. —0.35
15 —0.33	Radcliffe 2020.
Mar. 4 —0.21
18 —0.04	h. m. 8.
22 —0.18	1884, Mar. 22 7 47 5.39
Mean —0.222	25 5.86
B.J. 1884.0 24.996	Aug. 27 s.p. 6.02
h. m. 8.	Sept. 2 s. p. 5.89
Mean A. R. 1884.0 6 26 24.774	Mean A. R. 1884.0 7 47 5.790

OF ARTS AND SCIENCES.

MEAN RIGHT ASCENSIONS FOR ISSi.O. — Continued.

Gkoombkidge 1391.	SCHWEED	600.
h. m. 8.		h. m. 8.
1884, Mar. 15 8 1 59.07	1883, Mar. 22	9 57 18.01
18 58.63	Apr. 14	17.91
25 59.19	17	17.96
31 58.99	1884, Apr. 22	18.08
Mean A. R. 1884.0 8 1 58.970	23	17.83
	26	17.86
Radcliffe 2162.	Nov. 10 s. p.	18.12
li in s	21 s.p.	17.99
1884, Sept. 2 s. p. 8 3i 46.05 5 s.p. 45.61 9 s. p. 45.51 17 s.p. 45.93	24 s. p. Dec. 2, s. p.	18.03 18.08
Mean A. R. 1884.0	9 57 17.987*
Mean A. R. 1884.0 8 31 45.775	Radcliffe	2407.
Groombridge 1463.	1884, Oct. 24 s. p.	h. m. 8. 9 58 33.37
h m s	Nov. 10 s. p.	34.45
1883, Apr. 13 8 38 25.06 14 25.75	21 s.p.	32.82
Mean A. R. 1884.0	9 58 33.547
Oct. 5 s.p. 25.47 1884, Mar. 25 25.21
Radcliffe 2459. |
31 25.29		h. m. s.
Sept. 9 s. p. 25.04	1883, Oct. 17 s. p.	10 9 30.44
12 s.p. 25.45	1884, Apr. 22	29.88
17 s.p. 25.72	25	29.54
18 s. p. 25.72	26	29.83
Mean A. R. 1884.0 8 38 25.412	Nov. 10 s. p.	30.07
	21 s.-p.	29.86
Radcliffe 2189.	24 s. p.	29.86
h. m. 8. 1884, Sept. 5 s. p. 8 42 23.62	Dec. 2 s. p.	29.54
Mean A. R. 1884.0	10 9 29.878
9 s. p. 23.58 12 s. p. 23.92	Bradley	1458.
17 s.p. 24.21		h. m. 8.
Mean A. R. 1884.0 8 42 23.832	1883, Dec. 6 s. p.	10 31 58.49
	1884, Apr. 23	68.43
Radcliffe 2368.	26	68.80
	80	58.43
1884, Nov. 10 s. p. 9 39 9^86 21 s.p. 9.85 22 s. p. 10.39 24 s. p. 9.61 Mean A. R. 1884.0 9 39 9.928	May 2 Nov. 8 s.p. Dec. 2 s. p. 4 s.p. Mean A. R. 1884.0	88.61 58.75 58.74 58.70 10 31 58.619
	Radcliffe 2612. |
Schwerd 587.		h. na. s.
h. m. s.	1884, Apr. 26	11 0 19.94
1884, Apr. 22 9 44 48.90	May 2	19..37
23 48.82	3	18.91
26 49.39	Nov. 14 s. p.	19.20
Nov. 10 s. p. 49.43	15 s. p.	18.89
Mean A. R. 1884.0 9 44 49.135	Mean A. R. 1884.0	11 0 19.262
* A single observation of 1884 Nov. 22 s. p	. gives 1 s. more than the mean of these ten ;
it was rejected, because at the time .a workman	was hammering loudly in the observing room.
and a miscount of 1 s. may easily have taken j	ilace.

PROCEEDINGS OF THE AMERICAN ACADEMY

MEAN RIGHT ASCENSIONS FOR 188i.O. — Continued.

SCHWERD	665.	Groombridge 1845.
1884, May 2 Nov. 8 s. p. 14 s. p. 15 s. p. Mean A. R. 1884.0	h. m. s. 11 0 43.96 44.23 43.50 43.83 11 0 43.880	1883, Oct. 20 s. p. li 54 1-3.73 Nov. 21 s.p. 13.11 1884, Apr. 30 13.44 May 2 13.71 3 13.58 24 13.97 Nov. 21 s. p. 14.08 22 8. p. 13.57
Radcliffe 2684. |	29 s. p. 13.63
		Dec. 25 s. p. 13.62
	h. m. s.	Mean A. R. 1884.0 11 54 13.644
1884, Nov. 8 s. p.	11 22 52.02
14 s. p.	51.85	Groombridge 1850.
15 s. p.	52.10
21 s. p.	52.50	h. m. 8.
Mean A. R. 1884.0	11 22 52.118	1884, Apr. 30 11 58 52.71 May 2 53.57 3 53.70 24 54.07
Groombkidge 1782.	28 54.15
		Nov. 14 s. p. 52.66*
-	h. m. 8.	21 s.p. 53.71
1884, Apr. 23	11 23 37.02	Mean A. R. 1884.0 11 58 53.510
26	37.31
May 2 3	37.03 36.94	Groombridge 1858 := Bradley 3241.
Nov. 14 s. p.	36.80	h. m. 8.
15 s. p.	36.84	1884, Nov. 14 s. p. 12 5 45.30
21 s. p.	37.02	15 s. p. 45.77
Dec. 29 s. p.	37.31	21 s.p. 45.72
Mean A. R. 1884.0	11 23 37.034	22 s. p. 45.53 29 s. p. 45.87 Mean A. R. 1884.0 12 5 45.638
Radcliffe	2705.	Bradley 1642 (following of two stars).
	h. m. 8.	h. m. 8.
1884, Apr. 26	11 26 44.91	1884, Nov. 29 s. p. 12 11 9.46
May 2	45.36	Dec. 2 s. p. 9.41
3	44.63	4 s.p. 9.58
Nov. 21 s. p.	45.06	29 s. p. 9.71
Dec. 30 s. p.	45.94	Mean A. R. 1884.0 12 11 9.540
Mean A. R. 1884.0	11 26 45.180	Bradley 1672.
Radcliffe 2788.	1884, Apr. 30 C des T. +l'.01
		May 2 +170
	h. m. 8.	3 +0.31
1884, Nov. 15 s. p.	11 38 51.44	24 +1.76
21 s.p.	51.85	Mean +1.195
29 s. p.	51.99	C. des T. 18.33
Dec. 26 s. p.	50,78	h. m. 8.
Mean A. R. 1884.0	11 38 51.615	Mean A. R. 1884.0 12 14 19.525
* A large political meeting was going oa	in the neighborhood, with great uproar.

OF ARTS AND SCIENCES.

MEAN EIGHT ASCENSIONS FOR 1884.0. — Con^muerf.

Groombridge 1879.	SCHVTERD 761.
1884, Nor. 8 s. p.	h. m. s. 12 15 57.31	1884, Nov. 14 s. p. 12 41 53.46
14 s. p.	56.82	15 s. p. 53.44 21 s.p. 53.86 22 s.p. 53.59 Mean A. R. 1884.0 12 41 53.588
21 s. p.	56.52
22 s.p.	56.62
Mean A. R. 1884.0	12 15 56.818
Gkoombridg	E 1909. h. ni. s. 12 30 36.45	Fedorenko 2201.
1882, Nov. 21 s. p.	h. m. s.
188.3, May 17	35.67	1884, Nov. 15 s. p. 12 58 24.65
1884, May 2	36.25	21 s.p. [^wt.J 25.12
3	36.39	Dec. 2 s. p. 24.56
24	36.-33	4 s. p. 25.03
28	36.35 '	Mean A. R. 1884.0 12 58 24.800
Nov. 8 s. p.	36.17
21 s.p.	36.61
22 s. p.	36.91	Groombridge 2006.
29 s.p.	36.45
Mean A. R. 1884.0	12 30 36.358	h. m. 8. 1884, June 7 13 6 57.42 Mean A. R. 1884.0 13 6 57.420
SCHWEED	750.
1884, Nov. 21 s. p.	h. m. s. 12 33 37.52	Groombridge 1977.
22 s. p.	38.21
29 s.p. Dec. 2 s.p.	37.58	h. m. s.
37.29	1884, May 28 13 11 24.01
Mean A. R. 1884.0	12 33 37.650	30 24.28 June 7 24.12 13 24.13
SCHVTERD 752. 1	Nov. 15 s. p. 24.34
		1885, Jan. 14 s. p. 24.27
1883, Nov. 7 s. p. Dec. 6 s. p.	h. m. s. 12 34 38.02 37.66	19 s.p. 24.37 Mean A. R. 1884.0 13 11 24.217
29 s. p.	87.72
1884, Nov. 8 s. p. Mean A. R. 1884.0	38.99
12 34 38.098	SCHWERD 806.
^		h. m. s.
Groombeidg	E 1923.	1884, June 7 13 42 23.28 13 23.58
	h. m. s.	17 23.61
1884, May 24	12 37 33.93	18 23.51
28	33.75	Mean A. R. 1884.0 13 42 23.495
Nov. 8 8. p.	34.11
14 s. p.	34.07
Mean A. R. 1884.0	12 37 83.965	Groombridge 2063.
Groombridge 1927.	h. m. s. 1884, June 7 13 45 41.92
	h. m. s.	13 42.09
1884, May 24	12 41 29.37	17 42.14
Nov. 8 s. D.	29.63	18 41.88
Mean A. R. 1884'0	12 41 29.500	Mean A. R. 1884.0 13 45 42.008

10

PROCEEDINGS OF THE AMERICAN ACADEMY

MEAN RIGHT ASCENSIONS FOR 188i.O.— Continued.

SCHWEED 812.	Groombridge 2190.
1884, June 7 13 50 5.18 13 4.86	1884, June 18 14 58 13.90 21 14.03
17 4.95	26 13.67
18 4.77 Mean A. R. 1884.0 13 50 4.y40	July 3 13.31 Dec. 30 s. p. 13.96
	Mean A. R. 1884.0 14 58 13.774
SCHWERD 817.
1884, June 13 13 52 35.99	Groombridge 2213.
17 86.26
18 36.74	h. m. 8.
1885, June 9 36.43	1884, June 18 15 3 27.30
Mean A. R. 1884.0 13 52 86.355	26 27.18
	Dec. 30 s. p. 26.89
Groombridge 2099.	Mean A. R. 1884.0 15 3 27.128
h. m. 8.
1884, June 7 14 1 21.82
13 20.73	Groombridge 2283.
17 21.32
18 21.77 Mean A. R. 1884.0 14 121.410	1884, June 18 C. des T. —1.00 26 —2.41
	July 7 —1.20
5 Urs^ Minoris.	9 —1.77
s.	Mean — 1.595
1884, June 7 A. E. —0.03	C. des T. 0.52
13 +0.04
17 +0.01 18 +0.18	Mean A. R. 1884.0 15 14 58.925
Mean +0.050
A. E. 46.958
h. m. s.	ScHWERD 916.
Mean A. R. 1884.0 14 27 47.008
	h. m. s.
SCHWERD 853.	1884, June 18 15 27 38.79
	26 38.47
h. m. s.	July 7 38.94 9 38.55
1882, Dec. 12 s. p. 14 36 52.09
1883, June 14 51.91	Mean A. R. 1884.0 15 27 38.088
25 52.39
1884, June 13 52.32
17 52.32
18 52.54	e Urs^ Minoris.
21 52.83
Mean A. R. 1884.0 14 36 52.343	h. m. s.
	1883, July 18 15 34 53.15
Groombridge 2210.	20 52.94 23 52.59
h. m. s.	1884, June 26 52.63
1884, June 18 14 52 41.11	July 7 52.67
26 40.94	9 52.71
July 10 41.43	10 ' 52.76
Mean A. R. 1884.0 14 52 41.160	Mean A. R. 1884.0 15 34 52.779

OF ARTS AND SCIENCES.

11

MEAN EIGHT ASCENSIONS FOR ISSW.— Continued.

Gkoombridge 2286.	SCHWEKD 963.
h. m. 8.	h. m. s.
1883, Jan. 18 s. p. 15 39 33.34	1884, July 7 16 8 37.87
Feb. 12 s. p. 33.20	9 38.24
1884, June 18 33 87	10 38.04
July 7 33.85	11 38.19
9 34.10	Mean A. R. 1884.0 16 8 38.085
10 33.83
Dec. 30 s. p. 34.11	20 IJRSiE Minoris.
1885, Jan. 8 s. p. 34.28
Mean A. R. 1884.0 15 39 33.822	h. m. s.
	1884, July 18 16 15 28.00
	21 27.88
	25 27.90
Radcliffe 3523.	26 27.95
1884, Jan. 12 s. p. 16 o' 35.09 26 s.p. [|wt.] 34.76	Mean A. R. 1884.0 16 15 27.932
7j Urs^ Minoris.
July 11 35.34 15 35.57	1884, July 7 A.E.(Suppl.)+0.01
18 35.02	9 +0.13
21 36.14	10 —0.12
1885, Jan. 14 s. p. 35.02 31 s.p. 34.52	11 +0.05
Mean +0.018
Mean A. R. 1884.0 16 0 35.211	A. E. 54.388 h in. s
	Mean A. R. 1884.0 16 20 54.406
Geoombridge 2320.	Groombridge 2347.
1883, July 20 A. E. +o'22 24 +0.38 1884, July 9 +0.38 10 +0.21 15 +0.33	h. m. 8. 1884, July 18 16 22 4.77 21 4.82 22 4.93 25 4.82 Mean A. R. 1884.0 16 22 4.835
22 +0.47
Mean +0.332	A Draconis.
A. E. 0.350	s.
	1884, July 7 B.J. +0.07
h. m. s.	9 +0 08
Mean A. R. 1884.0 16 6 0.682	10 —0.14
	11 +0.20
	Mean +0.052
	B. J. 12.737
19 URS.a; Minoris.	h. m. s.
a	Mean A. R. 1884.0 16 28 12.789
1884, July 7 B.J. +0.21
9 0.24	Groombridge 2372.
10 —0.14
11 +0.16 Mean —0.002 B. J. 8.574	1884, July 7 16 32 12 J7 9 12.85 10 12.49
h. m. 8.	11 12.91
Mean A. R. 1884.0 16 14 8.572	Mean A. R. 1884.0 16 32 12.755

12

PROCEEDINGS OF THE AMERICAN ACADEMY

MEAN RIGHT ASCENSIONS FOR ISSi.O. — Continued.

Groombkidge 2373.	Groombkidge 2476.
1884, July 7 B.J. +0.13 9 —0.03 10 —0.24 11 +0.28 Mean +0.035 B. J. 38.894	h. m. 8. 1884, July 21 17 34 54.39 22 54.29 28 54.38 30 54.40 Mean A. R. 1884.0 17 34 54.365
Mean A. R. 1884.0 16 35 38.929	(0 Draconis.
SCHWERD 992. h. m. s. 1884, July 18 16 44 30.07 21 30.05 22 29.90 25 30.00 Mean A. R. 1884.0 16 44 30.005	1884, Aug. 2 B. J. —0.07 14 +0.04 15 0.00 18 0.15 Mean —0.045 B. J. 37.893 h. m. s. Mean A. R. 1884.0 17 37 37.848
Groombkidge 2391.	Radcliffe 3798.
1884, July 18 16 48' ]6i)4 22 16.69 25 17.02 26 16.86 Mean A. R. 1884.0 16 48 16.878	1883, Feb. 13 s. p. 17 38 2677 Mar. 1 s. p. 25.05 5 s. p. 26.57 Aug. 14 27.15 21 26.99
Eadclifee 3685. 1884, July 18 17 T 48.44 21 48.85 22 48.91 25 48.75 Mean A. R. 1884.0 17 1 48.738	22 26.24 1884, July 18 26.98 21 26.80 25 26 57 28 26.07 Mean A. R. 1884.0 17 38 26.609
35 Draconis.
f Draconis. 1884, July 30 B. J. +o!23 Aug. 2 +0.14 14 +0.07 18 —0.11 Mean +0.082 B. J. 25.697	1884, July 28 B. J. — o!o3 30 +0.14 Aug. 16 0.08 18 —0.11 23 +0.24 Mean +0.032 B. J. 38.592 Mean A. R. 1884.0 17 54 38.624
ix, m, 8. Mean A. R. 1884.0 17 32 25.779 Radcliffe 3749.	Groombkidge 2548.
h. m. 8. 1884, July 21 17 31 35.45 22 35.49 25 35.71 28 35.70 Mean A. R. 1884.0 17 31 35.588	1883, Aug. 14 18 4' 21.44 21 21.89 1884, Aug. 2 21.75 15 22.14 18 20.60 Mean A. R. 1884.0 18 4 21.564

OP ARTS AND SCIENCES.

13

MEAN RIGHT ASCENSIONS FOR ISSi.O. — Concluded.

Radcliffe 3900.

1884, July 30 Aug, 2

14

15 Mean A. R. 1884.0

h.

18

m. 10

18 10

8 03 7.54 7.66

7.98 7.802

Radcliffe 3903.

1884, Aug. 16

18

20

23 Mean A. R. 1884.0

h. 18

m. 11

18 11

14.79 14.38 15.22 14.87 14.815

Radcliffe 3921.

1883, Aug. 14

1884, Aug. 16

18 19 Mean A. R. 1884.0

h.

18

m. 11

18 11

3.81 4.46 3.74 4.13 4.035

(f> Draconis.

1884, Aug. 28

Sept. 1

2

4

B.J.

Mean B.J.

h. m

—0.02 +0.03 —0.05 +0.11 +0.018 25.235

Mean A. R. 1884.0 18 22 25.253

SCHWERD 1094.

1883, Aug. 14

1884, Aug. 19

20 23

27

[* wt]

h. m. s.

18 24 57.55 57.76 57.89 58.02 57.90

Mean A. R. 1884.0 18 24 57.802

Groombridge 2655.

1883, Aug. 14

1884, July 30 Aug. 14

15 16

B.J.

Mean B.J.

h.

—0.12 +0.07 —0.13

+0.11 0.00 v-0.014 20.928

Mean A. R. 1884.0 18 35 20.914

Groombridge 2708.

1884, Aug. 18

19

20

27 Mean A. R. 1884.0

h.

18

m.

38

18 38

36.47 36.73 37.06 36.23 36.622

Radcliffe 4165.

1884, Aug. 18

27

28 Sept. 2 Mean A. R. 1884.0

h. m. 18 53

18

48.33 48.20 48.56 48.22 53 48.328

Radcliffe 4253.

1884, Feb. 15 s. p.

Mar. 4 s. p.

25 s. p.

Mean A. R. 1884.0

h.

19

19

m. s.

6 19.06 19.67 20.16

6 19.630

SCHWERD 1139.

h.

19

1884, Mar. 22 s. p.

25 s. p. Aug. 18

19

27

28 Mean A. R. 1884.0 19 6 21.957

m. s.

6 21.59 22.15 21.97 22.00 21.71 22.32

Radcliffe 4313.

1884, Feb. 15 s. p. Mar. 4 s. p.

22 s. p.

25 s. p. Mean A. R. 1884.0

h. 19

m. s.

16 43.35

44.08

44.17

43.83

19 16 43.858

Cephei 39 H. s. p.

1884, Apr.

22 23 26 30

C. des T.

Mean C. des T. h. ni.

—0.03 +0.07 —0.31 —0.60 —0.218 50 26 s.

Mean A. R. 1884.0 23 27 50.042

14

PROCEEDINGS OF THE AMERICAN ACADEMY

II.

THUNDER-STORMS IN NEW ENGLAND IN THE SUMMER OF 1885.

By William Morris Davis.

Presented June 16, 1886.

Page

Collection of Material 14

Reduction of Observations 16

Daily Maximum Temperatures .... 16 General Account of Storms through the

Summer 17-38

Thunder-storms of July 9th and 21st . 38-54

Page

Review 64-58

Distribution of Storms through the Day . 55 Control of Thunder-storms by the larger

Cyclonic Circulation 56

Direction of Motion of Thunder-storms . 57

The observation of thunder-storms was taken as a special subject of investigation by tlie New England Meteorological Society during the summer of 1885. The considerable expense of the work was les- sened by effective assistance from the Chief Signal Officer, U. S. Army, and the remainder was borne by a grant from the Bache fund of the National Academy. By taking advantage of all available opportuni- tieis, offers of aid were received from about five hundred volunteers in various parts of New England, and about three hundi'ed of these main- tained their interest in the work through the summer, and sent in their records with considerable regularity. The Society is greatly indebted to these observers for their persevering assistance. As success in the work depended entirely on the enlistment of a well-distributed body of observers, effort was made to secure an equal representation of different districts, but not with as much success as it is to be hoped will attend a second season's campaign. Several districts, such as northeastern Connecticut and southeastern Massachusetts, were almost without ob- servers ; while only central and eastern Massachusetts were pretty well covered with stations. The results here stated are therefore in many ways to be regarded as preliminary to those that a more complete in- vestigation may yield in the current year.

The correspondence with volunteer observers during the summer months, the care of the records as received, and the preliminary tabu- lation and charting of the observations, have been in the hands of Mr. 0. N. Oswell, of the U. S. Signal Service, whose careful work has

OF AKTS AND SCIENCES. 15

been, from the beginning of this investigation, an essential in its con- tinuance. Mr. H. Helm Clayton, now observer at the Blue Hill Meteorological Observatory, has also given valued assistance in chart- ing certain storms. But even with so much effective help, a long delay in presenting this statement of our last summer's work has been un- avoidable. The discouraging effects of the delay may, however, be more than counterbalanced, if the report reaches our observers at the opening of another thunder-storm season, and, by pointing out the value of the work done last year, encourages them to still further efforts in a second campaign. These efforts should be of two kinds. First, the enlistment of new volunteers in neighboring towns ; it is hoped that our observers will take care — will indeed take pride in seeing that no large district remains unrepresented on our lists. Second, in giving special attention to certain questions which are mentioned in the following pages as needing well-directed, discriminate observation for their deci:?ion ; among these, the growth and motions of the storm-clouds, whether observed near at hand or on the distant horizon, may be named as of the first importance.

It is worth while to state here that one of the most serious difficuL ties that we have had to contend with is the lack of belief in the im- portance of trivial observations. Many observers have expressed some misgivings as to the value of their work, because they had only light rain and faint thunder to report ; and some persons seem really to have been discouraged even to the point of giving up their records, because they were not favored with any violent storms. It would take but a short time to convince all these observers that very simple records in many cases possess the especial interest of marking the beginning, end, or border of violent storms, without which it would be impossible to define the storm area or tract. Even the simple and direct statement of iVo storm to-day is important, as, for example, on July 3, when many small storms developed in eastern Massachusetts and moved out to sea. By consulting the map for that day, it will be perceived how important the report of " Distant thunder, but no rain," may become, when com- bined with observations at other stations. It is therefore especially desirable that observers should in all ca,ses follow instructions as closely as possible, lohether their record seems important or not. The impor- tance of an observation appears when it is brought together with all the other records of the same day. It is probably for similar reasons that many persons failed to see the necessity of timing their observations accurately ; a record that gives time only in round numbers, such as 3.50, 4.10, 4.30, is at once open to suspicion ; while one in which un-

16 PROCEEDINGS OP THE AMERICAN ACADEMY

even times occur with natural frequency, such as wind-squall 16.37, t:ain 16.39, heaviest rain 16.45 to 16.58 (this being an actual record from one of our best observers), bears strong internal testimony to its accuracy. In many storms, an inaccuracy of seven minutes in timing the beginning of rain would cause an error of almost six miles in charting the rain-front. An accurate timepiece, carefully recorded, is therefore essential to the best success.

Reduction of Observations. — On receiving records from our observ- ers, the forms were classified, first by dates, next by storms under the same date if possible, and last by States. Next, the time of rain be- ginning was charted, on a large map (six miles to the inch) if the records are numerous ; on an intermediate scale (seventeen miles to, an inch) for many dates ; on a small map (thirty-five miles to an inch) for days when the records were scattering. The greatest part of this work has been done by Mr. Oswell. When reduced to this form, the rain times are examined to discover evidence of the progression of the rain area across the country, and, if this appear clearly enough, lines indi- cating the probable position of the rain-front are drawn for the even hours,* or for smaller intervals, and from these the direction and ve- locity of the storm's advance are measured. Only a small portion of the work thus performed can be presented in this report. Having obtained a general idea of the stoi'm, the records are examined again in order of rain times, and the questions of wind directions, cloud motions, and temperature changes are considered.

Our insufficient number of stations has been a continual regret dur- ing this work ; the storms have in several cases seemed to choose districts freest from observers, so that in one sense the labor of this first season's investigation must be regarded as having its best result in exciting more general interest and better work another year. The storms of New England are also generally so moderate that the detec- tion of their circulation will be probably more difficult than if these studies were carried on in the Western Stajtes. Nevertheless, it is be- lieved that the collection of facts here briefly summarized will prove of value in the study of local storms in this country, and that it will serve well as the basis for more extended study in New England during the present year.

Daily Maximrim Temperatures. — The occurrence of thunder-storms is in many cases so closely connected with the change from high to low

* Throughout this report, the hours are counted from midnight to midnight, 80 that afternoon and evening hours fall between 12 and 24.

OF ARTS AND SCIENCES. 17

temperature that it has been thought advisable to present here a table, compiled from the records furnished for the Monthly Bulletin of the New England Meteorological Society, giving the mean of the daily maximum temperature for several stations through the summer. The stations chosen were all at some distance from the sea-coast, so as to be beyond the reach of the diurnal sea-breeze ; they were Brat- tleborough, Vt. ; Concord, N. H. ; Amherst, Princeton, Lowell, and Taunton, Mass. ; and Collinsville, Conn. ; all the records being from self-registering thei'mometers.

TABLE I.

Date. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

June 73 75 71 70 62 71 73 80 63 74 76 79 79 84 78 88

July 64 74 75 80 82 82 80 87 91 82 78 80 78 69 80 83

Aug. 85 75 72 84 83 74 74 79 77 78 84 86 85 85 73 75

Date. 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

June 75 78 82 85 79 77 69 77 84 83 82 81 80 68 —

July 92 86 82 87 90 86 84 82 88 87 83 84 82 75 85

Aug. 78 82 81 72 74 83 76 82 67 63 64 69 70 63 76

It is to the figures of this table that reference is made in the follow- ing general account of our thunder-storms, under the abbreviation " mean max."

General Account of Thunder-storms in the Summer Months.

June 1. Several thunder-storms appeared on this, the first day of systematic observations by our volunteers, and a brief account of them was issued in Bulletin No. 7 of the N. E. Meteorological Society ; they were close to the south of a centre of low pressure that crossed from Lake Champlain to the Gulf of Maine during that day, and the best defined storm moved eastward across southern New Hampshire at a rate of 31 miles an hour.

June 2, 3. During the approach and passage of an area of high pressure on these days, there were a few reports of light thunder- showers, of thunder heard, or of lightning seen ; but no general storm occurred.

June 4, 5. In the evening of the 4th, and during the 5th, observa- tions increased in number ; at this time, an area of low pressure was, according to the June Weather Review, advancing from southern Mich- igan to the New Jersey coast. On the 4th, a roughly defined storm of moderate strength, with rather heavy rain, advanced from south-

VOL. XXII. (n. S. XIV.) 2

18 PROCEEDINGS OF THE AMERICAN ACADEMY

western Connecticut towards central Rhode Island, from 20 J'' to 23^at about 40 miles an hour. It was the first of the season for this district. On the 5th, there was a general rain over New England, with low tem- perature and northeasterly winds during the morning, while the centre of low pressure moved from eastern New York southeastward over Long Island, or perhaps a little farther south. The rain seems to have bef^un, or at least to have distinctly increased, about the time that sev- eral peals of thunder were heard, their place of occurrence advancing southeastward with much regularity from southern Vermont at 5|^ to southeastern Massachusetts at 10\ with a velocity of 28 miles an hour. Several careful observers in Connecticut and Rhode Island report that the low scud clouds moved with the surface wind from the northeast, while higher clouds could be sometimes seen above, moving from the northwest or west, in accordance with the advance of the storm ; this was especially visible during a break in the clouds reported from sev- eral stations, at 8''.4o in central Connecticut, at 9\35 in the eastern part of the State, and about 10^ in Rhode Island ; but this cannot be considered as the "eye" of the passing cyclonic storm, as the barom- eter continued to fall for several hours afterwards, and the wind did not veer to north or northwest till noon or later, the temperature being from 60° to 70° during the morning. The time of passage of the lowest atmospheric pressure was a little after noon in eastern Con- necticut and Rhode Island ; it was accompanied by brisk northeast to north winds, thunder with an increase of rain, and a marked fall of temperature to below 50° in the later afternoon and evening, as the cloud cleared away with north or northwest winds. The afternoon thunder-storm moved eastward 20 or more miles an hour : it was not felt north of Rhode Island.

June G, 7. On June 6, there were no reports of thunder, (if we except an isolated report, probably misdated for June 5,) the signal service weather maps showing that New England was on this day tra- versed by an area of high pressure, moving southeasterly. On June 7, there was rising temperature with southerly winds, flowing in obedi- ence to the call of a low pressure centre north of the lower Lakes ; the day was fair, and no storms were reported till evening, when three observers, far apart in western, central, and eastern Massachusetts, report local or distant thunder-showers : at this time the centre of low pressure was nearing Montreal, on its way down the St. Lawrence.

June 8. This was a day of greater warmth (mean maximum, 80°, Table I.) and numerous reports, while the centre of low pressure tra- versed the Gulf of St. Lawrence. Thunder was heard and rain came

OF ARTS AND SCIENCES. 19

in showers with moderate or light southwest winds, sometimes veering to northwest shortly afcer midnight and during the early morning at many points ; a tolerably well-marked progression is found from central Vermont at 2^ central New Hampshire at 3^ to 4^ south- eastern New Hampshire and eastern Massachusetts from 4^* to G*"; but as additional showers are confused with the main one, and as the time of occurrence was unflivorable to reporting, little can be said of the storm. It is noteworthy, however, that a good number of observers made detailed records even at three and four o'clock in the mornine:. The rain was hardly felt south of AVorcester and Boston, although thunder was heard some distance farther south.

The temperature during the early storm was between 60° and 70''. The morning was fair, with southwesterly winds, and the warmth in- creased to 80° or 85° at noon. Clouds then began to gather, and be- tween 13'' and 14'^ developed into a long, narrow thunder-storm, stretching over one hundred and fifty miles, from northwestern Con- necticut to Cape Ann in northeastern Massachusetts. The intensity of the storm varied along this line, as seems to be generally the case, but was nowhere very severe ; at some stations, all the elements of the storm were moderate; at others, the rain, beginning gently, became heavy for a short time, and the heavy shower was preceded by a short- lived high wind from the northwest, and accompanied by a rapid fall of ten or fifteen degrees in the temperature. The electric action was inconsiderable at all stations. Northern Massachusetts and the States beyond had practically no thunder-storms in the afternoon. Starting from the beginning here defined, this storm extended southeastward, and reached eastern Connecticut, Rhode Island, and southeastern Mas- sachusetts about 16^ but was nowhere reported on the southern coast. The parallel lines representing its rain-front at 14^ lo^ and IQ^ are distant only about 20 miles, if measured N. N. W. to S. S. E. at right angles to their length ; but as nearly all observers agree in giving the storm a decided easterly motion, it would seem that the velocity should be measured very obliquely to the rain-front, in which case it may have been as much as 40 or more miles an hour. Although present- ing the well-marked linear extension characteristic of thunder-squalls, the clouds of this storm cannot be well classified from the records that we have other than as heavy rain clouds. The several types described under the storm of July 9 do not appear here. The observer at Provincetown, Mass., reports high stratus clouds moving from the west-southwest before the lower cumulo-stratus clouds came with the storm from the northwest.

20 PROCEEDINGS OF THE AMERICAN ACADEMY

June 9-15, During the first three days of this period, the morning weather maps show that an area of high pressure came from the northwest, and crossed New England; as its centre lay off our coast, an area of low pressure stood in Colorado. From the 11th to the 15th, this low pressure centre moved along the usual east-northeast path, and ran down the northern slope of the St. Lawrence valley ; during the latter part of its passage, there were three scattered reports of distant lightning in the evening of the 14th after a warm day (mean max. 84°). On the 15th, the day opened with low pressure centres in Newfoundland and Lake Superior, and high pressure off the Carolinas ; it was probably within the circuit of the Western area of low pressure that two local storms occurred, one in Maine in the afternoon, the other in Connecticut in the evening of this day. They brought us only three reports, in spite of the rather high temper- ature, and the not great distance of the low pressure area.

June 16. Several small showers occurred in southern New Eng- land while the second of the above-mentioned low pressure areas hesitated in the Gulf of St. Lawrence on the hot afternoon and even- ing of the 16th (mean max. 88°). The most interesting of these was a well-developed storm of small size that was felt a little after 14*' over the northern part of Rhode Island, and moved eastward to Cape Cod Bay. It was unfortunate in selecting a district poorly represented by our volunteers, and thus lost the opportunity of being carefully portrayed. It arose in northeastern Connecticut, and moved eastward across northern Rhode Island and southeastern Massachu- setts at a rate of about 35 miles an hour. The clouds were heavy, but the rain-fall was generally moderate ; the temperature fell from 90° or 95°, as it had been shortly after noon, to 80° or 75°, and rose again a little after the storm moved past ; the wind showed a distinct control by the storm, as a brief gust of moderate strength was felt at most of the stations jnst before or during the rain, the direction of its blowing being in all cases away from the centre of the rain area. On reaching Taunton, the rain had become very light ; and no report of the storm came from Plymouth or Provincetown. Jewett City, Conn., Crompton, R. I., Fall River and East Freetown, Mass., saw the storm passing to the north ; and Medfield and Pembroke, Mass., reported it to the south, about the time another storm was passino- to the north. This second storm formed not far north of the first one, at a little later hour, and passed over Blue Hill on its way to the south shore of the Bay. The rain was reported heavy only on Blue Hill, where the wind was strong for a time during the shower.

OP ARTS AND SCIENCES. 21

This storm was visible from Cambridge as a great mass of boilin<r cumulus clouds overlain with fiat, outspreading cirrus; their angular altitude was estimated at 30°, giving a height of five or six miles. A third storm, or group of showers, less defined than the others, appeared first on the Connecticut River in Massachusetts about \b^ to 1 &", and gave moderate showers to various stations on its way to the coast, at and north of Boston, where it arrived at 21'' or later. The first two storms here described were much like, but smaller than the noon thuuder-squall of July 21, that is described in detail farther on. Very numerous and careful observations are needed for the full definition of such storms.

June 17-21. Another period without thunder-storms now appears. On the morning of the 17th, a large area of high pressure appeared over the Mississippi valley; on the 18th, it was on the Ohio valley ; on the 19th and 20th, on the Middle Atlantic coast. On the 21st, an ill-defined area of low pressure was central on Lake Michigan. All these days were fair and warm in New England, and on the 20th the mean maximum temjierature was 85°.

June 22. The low pressure area was central north of Montreal in the morning, with barometer down to 29.40 ; the isobars were strongly V-shaped, with the axis extending down the Hudson valley ; but in spite of this apparently favorable condition for the collision of warm southwest and cool northwest winds, New England had only light showers, and among these there were several without thunder. No well-developed thunder-storm appeared. The temperature was mod- erate (mean max. 77°).

June 23-25. This period is a repetition of that from June 17 to 21 ; an anticyclonic area moving from the Mississippi valley to the- sea-coast, with fair weather in New England ; the mean maximum temperature was only 69° on the 23d, but rose to 84° on the 25th.

June 26. A number of stations in southern New Hampshire, and a few in southwestern Maine, report moderate thunder-storms on this date at noon, in the afternoon, and at night. Stations were wanting farther north, so it cannot be said how far the storms extended in that direction ; but they were definitely absent in the south, in spite of continued warm weather (mean max. 83°). These storms are less distinctly dependent on cyclonic conditions than any of their pre- decessors, as there was no well-defined low pressure centre within the limits of the weather map at the time; a high pressure area was mov- ing eastward from the Middle Atlantic States.

June 27. An area of slightly diminished pressure appeared on this

22 PROCEEDINGS OF THE AMERICAN ACADEMY

morning north of Lake Superior, and moved to western Canada in the course of the next twenty-four hours. During the afternoon of this hot day (mean max. 82°), several rather strong storms occurred in northeastern Massachusetts, and scattered reports came from farther north and west. Showers without thunder are also reported. The two storms that can be partly defined began in northeastern Massachu- setts ; the first about noon, the second a little farther west than the first, at IS*" to 14^ Both moved east-northeast at about 20 miles an hour, and passed out to sea over Cape Ann and farther north. On account of their coming so close together, it is difficult to distinguish them in some of the reports, but they showed the usual control over the winds and temperature at a number of stations. At a moderate distance away, the wind blew with moderate or light velocity towards the storm ; nearer the rain area, and especially at times of heavy rain, the wind blew away from it, and for a time with increased strength. North of Salem, where the rain was especially heavy, the wind blew outward even at the back of the storm ; this seldom appears, and in this case seems to be connected with the low velocity of the storm, and with its lateral or backward growth to the westward, even while mov- ing east. The storms as a whole were seen from Cohasset, south of the Bay, moving slowly in an easterly direction ; their angular altitude was estimated to be 20° ; their distance must have been at least 20 miles, and was more likely 22 miles. This would make their height six or more miles.

An appreciative record from Southampton, Mass., on this date, describes in detail the extension of the rain area in a local thunder- shower in a direction contrary to the movement of the clouds, as above suggested. Cumulus clouds of ordinary type had been seen growing to larger and larger size from 10'\30, and moving from the southwest while the surface breeze blew towards them from the southeast. About noon, rain was seen falling from the clouds, and thunder was heard a little later. Then additional cumuli, cominff from the south- west, developed very rapidly as they approached the shower, build- ing the thunder-cloud southward. " The shower as a whole had a movement to the northeast, but the development of the cloud area in the manner noticed exceeded the velocity of the shower movement." Other observers describe this shower as moving slowly ; it would be an excellent one for study had our stations been more numerous in this district. Its motion and that of several neighboring showers was from west to east, or a little to north of east, and the velocity seems to have been moderate, although this is poorly defined by our records.

OF ARTS AND SCIENCES. 23

June 28. The low pressure area of the preceding day moved east- ward with warm weather, and while central over Lake Ontario, on the afternoon of the 28th, a moderate number of showers were reported in New England, generally light and in most cases moving from the west or southwest. Towards night, the thunder died away, but the rain increased.

June 29. (Storm charted by Mr. H. H. Clayton. See Fig. 1.) No better example than the storm of June 29th can be taken to show the need of some better classification of storms than that based simply on the fact of thunder being heard. The storm of this date was one of a series of rain showers that accompanied the approach and passage of the area of low pressure already mentioned. The night before (June 28), when the centre of low pressure was in northern New York, there was a general rain at many stations, but only a few re- ports mentioned thunder ; the morning of the 29th was cloudy, and the air rather close and oppressive, although the thermometer did not often rise over 85° to 90° (mean max. 80°). The centre of low pressure, according to the Monthly Weather Review, moved south- eastward during the day, from northern Vermont to Massachusetts Bay, and the next morning (June 30) lay in the Gulf of Maine. The afternoon thunder-storm began in or a little west of the Connecticut valley, just before noon, and moved about east or east-southeast, eighteen to twenty-five miles an hour, fading away as it approached the eastern and southern coast, before sunset. Its front was not well marked ; the rain began gently, and there was seldom any clearly marked previous change in the wind's direction or strength, though, at some stations where the rain became heavy, the wind then shifted for a time to northwest and became '' brisk"; nor was the rain-front con- tinuous, as central Massachusetts had no rain, while Connecticut, eastern Massachusetts (about the Newtons), and southern New Hamp- shire had heavy rain. In Connecticut, the wind before the rain was light, from south or southwest, with moderately high temperatures (80°-85°) ; it shifted during the shower to west or northwest, and the temperature fell fifteen or more degrees (68° to 70°). At Hartford, Middletown, and Colchester, the rain was heavy, and hail as large as robin's es:G;s fell for a few minutes at the latter station.

The New Hampshire portion of the storm that came at the same time as that in Connecticut had less defined characteristics ; the rain began very irregularly, and the changes of the wind seem to be with- out system, some stations reporting a change from southeast to north- west, others almost the reverse, and still others had a variable

24

PROCEEDINGS OP THE AMERICAN ACADEMY

succession of light breezes. Nearly all stations agree in giving the storm an advance from some westerly point, and, according to the times of rain beginning, its progress was at about the same rate as in Connecticut, in spite of its standing to the north of the neighboring centre of low pressure.* About Concord, the rain was very heavy.

The northern and southern ends of the storm seem to have faded away while the central part increased in strength as it approached the eastern coast. There was heavy rain and some hail near Haverhill, Mass., and all the many Newton stations, along with Boston and those near by, give heavy rain ; but the most marked feature of the storm

. \ . / I

!s3

Fig. 1.

was the prevalence of easterly winds charged with low, foggy clouds, just before and during most of the rain ; this was sometimes inter- rupted by a "moderate" or "brisk" northwesterly breeze about the time of the heaviest rain, but the east wind generally returned as the rain decreased. The indraft towards the storm in front of the rain, and the temporary control of the wind's direction by heavy showers, here find good illustration. Lightning strokes seem to follow the

* In Bavaria and Austria, it has been found that thunder-storms in this position very commonly move to the west.

OP ARTS AND SCIENCES. 25

heavy rain ; four are reported about Waltham and Newtonville, others in Concord, N. H,, Millville, Mass., with hail, and Middletown, Conn.

A composite portrait (see special description of squall of July 21) of this storm (June 29), prepared by Mr, Clayton, brings out with much distinctness its difference from the thunder-squalls of July 9th and 21st, in the absence of strong out-blowing winds in front of the rain, and in the presence of a district indraft of eastei-ly winds in front of the middle and northern half of the storm. Durinjj the rain, no safe generalization can be made concerning the wind direc- tions ; it should not be inferred from this that there is no definiteness in the circulation of the wind at that time, but rather that this method fails to discover it. With more numerous stations, and more regular observations, there is good probability of discovering system where only disorder now appears.

June 30 and July 1. The centre of low pressure above described lay over the Gulf of Maine on June 30, giving us cool weather, and only three records of thunder were received for this day. On the 1st of July, the centre moved northward across western Maine, with still cooler weather (mean max. 64°), and no reports of local storms were made: they may have occurred in eastern Maine, where we had no observers. The abrupt change in the character of our weather on the two sides of a centre of low pressure is thus finely illustrated.

July 2. During the morning hours the centre of low pressure retrograded into western Canada, giving us warmer weather again ; in the afternoon and night it returned across northern Vermont, and while there a moderate storm arose, about IT'^ to 19'^, in central Massachusetts, and moved east-northeast, becoming rather strong at Lancaster and Sterling at 21'', when the rain was heavy, and two buildings were struck by lightning. The rain extended little farther. The lightning from this storm was visible from a number of stations on the northeastern coast at the above-named hour. There were a few other isolated showers in the evening.

July 3. The same centre of low pressure still controlled our weather, moving on this date from Mt. Washington north to the St. Lawrence valley ; and a number of separate storms appeared in the afternoon in central and southern New Hampshire and central and eastern Massachusetts. These local storms followed one another in such close succession in eastern Massachusetts that it has been impossible to trace them all. Ten of the more distinct ones are rep- resented in Figure 2, moving southeast or east-southeast, at different hours, their velocity averaging sixteen miles an hour. In several

26

PROCEEDINGS OF THE AMERICAN ACADEMY

of the storms the clouds seem to have been simply overgrown cumuli; in others the ordinary high-level cirrus overflow is mentioned. An

Fig. 2.

inflow of winds before or alongside of the storms is seen in several cases, followed by a reversal or deflection of the wind as the area of heavy rain approaches.

July 4. The distribution of atmospheric pressure east of the Mississippi was rather uniform on this day, with gentle northward gradients and rising temperature. A few light local storms appeared in central New England in the late afternoon and evening.

July 5. The barometric gradients were on this day directed west- ward, from a high pressure area off the coast to a low pressure centre in Minnesota. The temperatures were high, but not excessive (mean max. 82°), and the day passed with trifling showers ; but in the evening, from 18^* to 22^ a very severe storm, with heavy rain and hail and incessant lightning, striking in many places, passed eastward across central and northern New Hampshire. Our observers there were too few to define it. Lightning from this storm was very noticeable south of Boston, over one hundred miles distant. In eastern New York a destructive hail-storm occurred near Chatham- Four-Corners at about 18'' ; its path was five miles long and one wide. No reports of it were received from western Massachusetts.

These two are the first strong storms of the summer that occurred without close connection with a centre of low pressure.

OF AETS AND SCIENCES. 27

July 6. The western centre of low pressure on this day followed a curious retrograde course from Lake Superior towards Lake Winni- peg. Under its influence, New England had warm southerly winds, turning to southeast or east along the eastern coast, and southwest inland. Several local showers of no great intensity arose in Vermont, New Hampshire, and Eastern Massachusetts in the afternoon and evening, moving east-northeast at a moderate rate ; they appeared to be developed from ordinary cumulus clouds, whose motion was per- sistent from the southwest or south, little affected by the inflection of the surface winds along the coast. As an example, a shower that passed over the south shore of Boston Bay may be mentioned. The morning sky had been nearly covered with light cirro-cumuli, beneath which large cumuli formed rapidly about noon, and united into a long nimbus stretching about north and south, and moving to the north- east : the light surface winds at one station were east before the rain, calm during the light shower, west for a short time after the shower, and then east again through a fair afternoon. Other stations report a continued east or southeast wind during and after the passage of the shower. The showers in New Hampshire were heavier: in the east- ern part of the State they also presented the peculiarity of advancing from the southwest over an easterly wind.

July 7. A few local showers, with moderate electric action. The winds were still warm southerly, under the influence of high pressure off the coast and the low pressure centre that hesitated over Lake Winnipeg.

July 8. A new centre of low pressure appears on the morning weather map for this date in Dakota. Violent thunder-storms appeared south of the Great Lakes, but none formed in New England, although the temperature rose rapidly (mean max. of 7th, 80°; of 8th, 87°).

July 9. The newly formed area of low pressure moved eastward, north of the lower lakes, and on the afternoon of this day stood near Quebec. Under its influence the mean maximum temperature rose to 91°, and the largest storm of the summer was developed, extending with almost continuous front from southwestern! to northeastern New England, and accompanied by all the characteristics of a full-grown storm. It forms the subject of a special description on ]). 38.

July 10. The low pressure centre of the previous day turned southward from the St. Lawrence valley, and this morning stood in western Maine. Had this attitude been preserved through the hotter hours of the day, we should doubtless have had more than scattered reports of light thunder-showers and heat-lightning; but it moved away eastward, leaving an area of cool fine weather behind it.

28 PROCEEDINGS OF THE AMERICAN ACADEMY

July 11 to 20. This peculiar period of ten days almost free from storms is oue of the most marked features of the summer. There were no reports on the 11th, 12th, or 13th, duriug the passage of an area of high pressure, and the approach of two low pressure centres which passed north and south of New England on the 14th. The northern of these cyclonic centres ran down the St. Lawrence valley, on a path seemingly well adapted to giving us local storms ; but the other one moved northeastward from New Jersey, probably along the southern coast of New England, and controlled our weather, giving us a light general rain, with easterly or southeasterly winds and a low temperature (mean max. 69°). A few reports of moderate thun- der came from Rhode Island in the evening. On the 15th, the tem- peratui-e rose, but there were no reports. The 16th brought an area of high pressure along the Atlantic coast, with still rising tempera- ture (mean max. 83°) ; a few sharp local storms were here developed, the most severe being in central New Hampshire at 15*^, where the lightning struck in several places ; but reports were received from only five observers. The absence of storms on the 17th is remarkable. A low pressure centre had come fi-om the West to Lake Pluron, and thence over Canada north of the St. Lawrence. It produced the hot- test mean maximum temperature of the summer (92°), with the char- acteristic circulation of winds that in other cases developed violent storms ; and yet on this day there were only three reports, all of heat- lightning in the evening. "We are led to suppose that the position of the low pressure centre and the arrangement of winds proper to give storms in New England was not reached until night, when the absence of sunshine prevented their development. Still, it is surprising that none came to us that had formed farther west in the daytime.

The ISth, 19th, and 20th furnished no reports; they were charac- terized by the passage of an area of high pressure, with fair, warm weather.

July 21. After this long period of inaction in New England, there came a low pressure centre that had been loitering in Dakota on the 18th and 19th, and that moved rapidly eastward on the 20th. The morning of the 21st found it north of Lake Huron, and at 15'* it was in southwestern Maine : its progressive velocity was therefore about 60 miles an hour during this interval. Several storms accompanied this low pressure area. Two of them are among the most interesting of the summer, both being notable for their high velocity of 48 and 43 . miles an hour, and their violent wind squall. They are described in some detail on p. 45. The mean maximum temperature for the day was 90°.

OF ARTS AND SCIENCES. 29

July 22, 23, 24. These days were occujiied with the passage of an area of high pressure, with fair, warm weather, and had only three, three, and seven reports respectively : the most severe of the last day was in central New Hampshire, where our lack of observers prevented its receiving more notice.

July 25. An area of low pressure moved from Lake Michigan on the 24th north of the St. Lawrence, and was north of the Gulf on the afternoon of the 25th, when several storms occurred in southern New England, after an oppressively hot day. The complication of successive storms at a number of stations makes it difficult to deter- mine their advance. The most distinct one moved from the Connecti- cut valley, at 1A\ east-northeast across Massachusetts, reaching the sea about 16^', and having a velocity of about 35 miles an hour: its rain-front stood northwest and southeast, a somewhat unusual atti- tude. Heavy storm-clouds were developed over southern central Connecticut about 15^^ to IG"*, but they gave only moderate rain. Few reports make mention of violent winds, although the rain-fall was not infrequently heavy. These storms did not mark any distinct change of weather, as the mean maximum temperatures for the 25th and 26th were 88° and 87°.

July 26. On this date there was still low pressure over the Gulf of St. Lawrence, and high pressure north of the Lakes. The day became oppressively hot, but only two thunder-storms of any defined form can be recognized. One crossed the Connecticut valley near Amherst, Mass., about noon, moving from northwest to southeast, with heavy rain preceded by high northwest wind, and cooling the air from 90° down to 80° ; it seems to have been short-lived, as stations farther east did not report it. The other and larger one bejran south of Framingham and Natick and north of Medfield and Norwood, about r2'\45, and moved southeast, crossing Martha's Vine- yard at 15'M0 to 15\40; its velocity therefore averaged about 23 miles an hour, but the time of the storm's arrival at intermediate stations implies a gain in velocity in the latter part of its course. The wind of the day was generally northwest ; it was not seriously affected by this storm till the cloud had passed Taunton (13'\50) and reached East Freetown (14''.14) ; here and beyond, the wind before the storm was light southerly, with temperature 92° inland, and 85° to 80° on the coast; then, with the beginning of the rain, the wind-squall came high from the northwest, and the temperature fell ten or fifteen degrees in half an hour or less. The rain became very heavy, and numerous lightning strokes are reported near New Bedford and on

30

PROCEEDINGS OF THE AMERICAN ACADEMY

Martha's Vineyard. On Vineyard Sound and on the island the storm was severe. At Cottage City, on the island, the storm-clouds were seen in the distant north at 12''.30, when the storm was forming, seventy or eighty miles away.

July 27, 28. The passage of another area of high pressure again brouo-ht fair weather, with only two reports of thunder on the latter date.

July 29. (Fig. 3.) During the previous day the baric gradients were very gentle, and our temperatures were high (mean max. 84°). On this morning a slight baric depression lay north of Quebec, and still maintained warm weather under its southwesterly winds (mean max. 82°). Five separate storms have been traced for this day, the charting being done by Mr. Clayton.

a. The first one came early in the morning, being reported in northern Vermont at 2 A. m., and reaching eastern New Hampshire about eight o'clock, having travelled over a narrow path to the east- southeast at a rate of perhaps eighteen miles an hour : it does not seem to have reached the sea-coast.

Fig. 3.

h. The second storm advanced eastward along the boundary of New Hampshire and Massachusetts, at a rate that may be estimated at twenty miles an hour, and passed out to sea at noon.

c. The third storm developed in northeastern Massachusetts about noon, and moved eastward at the same leisurely rate as its predecessors. Its rain-fall is recorded as over half an inch at several stations; its clouds were dark and heavy, and several lightning strokes are reported ; but

OP ARTS AND SCIENCES. 81

there was no squall connected with the storm, the wind remaining fresh southerly or southwesterly during the rain, except at Haverhill and Newburyport, on the northern side of the storm, where it was reversed to northeasterly.

d. The fourth storm probably consisted of several parts : it extended over central Connecticut in the middle afternoon, but its motion cannot be well determined for lack of observations. Fine festoon clouds were seen on the after side of its cirro-stratus cover in the southern part of the State.

e. The last and most extensive storm of the day began in southern New Hampshire about 14'\ and moved to the southeast about twenty- five miles an hour. It does not seem to have been a well united storm, but consisted of numerous loosely connected parts, from which showers of varying strength fell. At Lowell storms c and e together yielded about four inches of rain, one gauge reporting 3.93 and another 4.20 inches; and of this the greater part fell from 15\15 to 16^.15, without much wind, althougli the clouds were much agitated as the storm approached. Violent winds appeared only in a limited district north of Worcester, about Holden and Princeton, and hail fell a little farther north, at Westminster. -'

Mr. Clayton has constructed several synoptic charts for storms c and e, as far as the observations would allow, showing the extent of the rain area and the direction of the wind at certain even hours. Owing to the failure of many observers to carry out their instructions literally, it was found impossible to make these charts very complete; but the later ones show a tolerably distinct cyclonic motion of the winds within and around the oval rain area, implying that the thunder- storm area possessed a gentle spiral rotary circulation on a small scale, as has been determined for storms of this kind in Europe. On this date the very moderate gradients gave good opportunity for the forma- tion of such " secondary depressions." Especial effort will be made to accumulate data for such synoptic charts during the coming summer, as they promise to be of particular interest in connection with the smaller local storms.

July 30 possessed no well-defined centres of high or low pressure ; its temperature was lowered (mean max. 75°) by the passage eastward of the area of low pressure that controlled the weather of the preced- ing day. A few scattered reports were received, the only storm of any severity being in central New Hampshire about 8^.

July 31. This was a day of generally equable distribution of press- ure, of rather high temperature, and with winds of indeterminate circu-

32 PROCEEDINGS OF THE AMERICAN ACADEMY

lation. Through the afternoon, showers of local formation began at a number of places over New England, and moved about without coales- cing into any general storm. Their distribution seems to have been more detailed than that of our observers, so that it is quite impossible to follow their tracks. The rain was seldom heavy, but when so, there seems to have been a wind of more or less strength caused by it, blowing away from the storm ; in such cases, the breeze was often felt at points be- yond the area of the heavy rain. No definite statement can be made as to the direction or rate of the storms' motion, but a number of ob- servers speak of the cloud velocity as slow, and record the direction as W. to E., or N.W. to S. E. A marked peculiarity of the day was the severity of the lightning and thunder with but moderate rain-fall, and the extension of the area where thunder was heard twenty or thirty miles southeast of the margin of the rain area.

Auw. 1 had an even more equable distribution of pressure than the previous day, and its maximum temperatures averaged again 85°. Thunder-storms developed about noon, and from 16^ to 18'' a large storm of much severity was felt in central and eastern Massachusetts. The wind of the storm was generally light, but in central Massachu- setts showed the outblowing squall. A newspaper report mentions destructive wind (tornado ?) at Bernardston, Mass. The rain was very heavy at many stations, hail fell at a few places, and twenty -five strokes of lightning were reported, although attention was not given especially to this question. The early motion of the storm cannot be satisfacto- rily determined ; much time has been spent in attempting to analyze the many records gathered, but without good success, on account of the irregularity in the line of rain-front and the complicated suc- cession of showers. After raining heavily between Worcester and Boston, the storm moved southeast, with a velocity of twenty or thirty miles an hour, decreasing in intensity as it approached the coast after sunset.

Aug. 2. A low pressure centre now appears in Illinois, moving northeastward. Under the influence of its easterly winds, the tem- perature fell ten degrees below that of the day before, and rain was recorded at a number of points. Few thunder-storm reports were received ; several from southern Rhode Island imply a small storm of some severity, with heavy rain, moving eastward about noon.

Aug. 3. The low pressure centre, still moving slowly to the north- east, occupied this day in passing between Lakes Huron and Erie ; in New England, the temperature was moderate (72'^), with east- erly winds and rain ; very few stations reported any thunder. On

OP ARTS AND SCIENCES. 33

the afternoon of this day, several tornadoes occurred in eastern Pennsylvania.

Aug. 4. The low pressure centre passed slowly along north of the St. Lawrence, and gave rain that began gently on the afternoon of the 3d, and continued through the night, becoming very heavy, with strong south or southeast wind, just before clearing in the morning, and this latter part of the rain was accompanied with violent thunder and light- ning, striking in several places. Many observers speak of this storm as a peculiar one, describing it as an ordinary southeast storm during the night, but culminating in a violent thunder-storm with torrents of rain as it was about to break away. The time of rain beginning can- not be used in this case as a guide to the progress of the storm ; but the hour of loudest thunder and heaviest rain serves well instead. This shows the storm to have passed over western Connecticut and Massa- chusetts and southern Vermont shortly after midnight, reaching the Connecticut valley at 2'', Rhode Island to the upper Merrimack valley between 8'^ and 9^, Boston about 9^.30, Salem and Newburyport at lO'^, Cape Ann at 10^30, Provincetown at ll'', and appearing later on the Maine coast. This gives an east-northeast progression of ten or twelve miles an hour. The storm may have increased in intensity from central Massachusetts, where several good observers made no report, to the eastern part of the State, where reports are numerous, and where many stations measured two or three inches of rain. The wind was not noticeably affected by the rain ; it was strong south- east, with heavier gusts during the storm, and soon fell off to gentle southwest as the sky cleared. The temperature was rising in eastern Massachusetts before the heavy rain, and fell five or ten degrees dur- ing its continuance. As the storm passed away, the temperature rose rapidly, and the mean maximum of the day was 84°. This storm was therefore quite unusual in separating a period of cooler from a period of warmer weather.

A second storm was felt in Vermont in the late afternoon and even- ing, possibly derived from an afternoon storm in New York State ; its lightning was seen in eastern Massachusetts. Many observers also report heat-lightning in the southeast at 20*^ to 22^ but we have no direct record of the storm from which it came except that the observer at Provincetown saw a bank of low clouds out over the ocean at these hours, with frequent and brilliant flashes of lightning.

Aug. 5. A warm day, with southwest winds drawn towards the low pressure centre, now north of the Gulf of St. Lawrence. A thunder-storm came down the Mohawk valley shoi'tly after noon,

VOL. XXII. (n. 8. XIV.) 3

34 PROCEEDINGS OF THE AMERICAN ACADEMY

crossed the Hudson at Albany at 14''.25, traversed northern Massa- chusetts and southern Vermont and New Hampshire during the after- noon, moving eastward at about 27 miles an hour, and reaching the sea-coast about 19''.40, at the mouth of the Merrimack; its rain-fall was not heavy, although the lightning was active and the out-blowing wind was tolerably well marked in advance of the storm. A simi- lar storm traversed southern Massachusetts, Connecticut, and Rhode Island at the same rate about an hour behind the first one ; its clouds were heavy, the lightning was brilliant and struck in several places, and the change from warm southwest to cool, brisk northwest winds was generally distinct. The rain was heavy, but brief, until it de- creased and nearly disappeared on reaching Massachusetts Bay, about 20\30.

Aug. 6-11. This period was relatively cool and free from thunder- storms, with the exception of a light shower on the 11th, correspond- ing to a distinct rise in temperature on that date. The time was chiefly occupied with the passage of high pressures with low gradients, except on the 10th, when a vague depression, with very faint gradients on its western side, passed over Canada. The shower of the 11th seems to be connected with the warm southerly winds on the north- western side of a high pressure area.

Aug. 12-14, Following the high pressure above named, there came a low pressure centre over Lake Superior on the 12th, that moved eastward and became better developed on the loth. From the 6th to the 10th, the temperature was moderate (mean max. 74-79°) ; from the 11th to the 14th, it became distinctly warmer (mean max. 84-86°). On the 12th, there were ten reports that seem to belong to a sharp local storm moving from southwest to northeast across southern New Hamp- shire from 15'' to n** or 18**, with moderate rain and west or south- west squall. The morning of the 13th was generally cloudy, close and oppressive, — "dog-day weather." About noon, several local showers with southerly winds were reported in southern New England, but, as a rule, not near the coast ; in central New Hampshire near the moun- tains the rain was heavier. Later in the day, a shower or series of showers stretched from southwest to northeast over Connecticut and central Massachusetts, and moved obliquely to the east-northeast ; at this time the lower clouds came from the south, and the upper from the west. Rain was recorded at several stations without thunder, and seemed generally without the characteristics of a thunder-storm ; the day would have been called showery rather than stormy. On the morning of the 14th, the low pressure centre stood over the lower St.

OF ARTS AND SCIENCES. 35

Lawrence, and it was not till after a high temperature had been reached with southerly winds (mean max. 86°) that the cooler northwesterly winds appeared. It is rather curious to note that this change was not accompanied by distinct storms, for only six reports came in for this day, and of these only one (from Mayfield, Me.) describes a storm of any violence.

Aug. 15-17. There were no reports received for this cool period of high pressures and fine weather.

Aug. 18. During the approach of the succeeding area of low pressure and consequent rise of temperature under southwesterly winds, there came another period of storms. Several showers were felt on the 18th. A small one occurred in southern Rhode Island and Mas- sachusetts about sunrise. A much larger one arrived in northwestern Massachusetts about 16'\ and moved southeastward, at a rate of 20 to 30 miles an hour, into Rhode Island at 20'' ; its rain was rather heavy, and the lightning severe at several places, and there was generally a change from moderate southerly to gusty westerly winds as the storm approached. A third and less defined storm passed over nearly the same district in the evening. At this time, the low pressure area was central in Canada north of Montreal.

Aug. 19. Under the influence of the same low pressure area, which moved slowly northeastward, a high temperature was maintained into the 19th, and the day began with a storm that came from New York a little after midnight. It was felt as a heavy rain at Albany about one A. M. ; a few reports at early morning hours in Massachusetts and Connecticut probably refer to the same storm, and at 5-6^ it passed over Rhode Island. This would make its velocity and direction about the same as those of the storm the afternoon before, but this conclusion is not very securely based. In the afternoon, several showers occurred in Vermont, New Hampshire, and Maine, but they cannot be safely traced, owing to the want of sufficient observers. At about the same time, a storm of some strength passed over Connecticut and Rhode Island, moving east-southeast about 30 miles an hour. It entered northwestern Connecticut about 15''.30, with dark clouds, a brisk to high squall from the northwest, followed by heavy rain and some hail. At Norfolk, the temperature fell from 75° to 57° in an hour, and 0.87 of an inch of rain was collected in eighty minutes. But this degree of strength was not long continued, and in Rhode Island the storm faded away between 18'' and 19*^.

Aug, 20, 21. During the passage of an area of high pressure, the maximum temperature fell ten degrees below that of the preceding

36 PROCEEDINGS OF THE AMERICAN ACADEMY

period, and no storms * are recorded till late in the evening of the 21st.

Auw. 22. A low pressure area moved eastward north of Lake Huron on the afternoon of the 21st, and crossed over Canada in the nio-ht, its centre passing well to the north of the St. Lawrence. Sev- eral interesting storms occurred in connection with it. The first was the outgrowth of showers that began in the evening of the 21st, with southerly winds. At a number of stations, the rain began some time before thunder was heard, but increased greatly in amount when the thunder-storm was passing. Our first records are at 22'' on the even- ing of the 21st, from Burlington, Vt. ; then come a number of midnight or early morning records from southern Vermont, New Hampshire, and western Maine ; at 3^ to 4**, the storm had reached central Mas- sachusetts and southeastern New Hampshire ; about b^, it extended along the coast from Boston towards Cape Ann, and lightning was re- ported in the north from Provincetown at 6''. Southern New Eng- land did not feel this storm. It may therefore be concluded that its motion was easterly, and at a rate of about 35 miles an hour, while its front stretched obliquely from southwest to northeast. Although oc- currino' at hours inconvenient for observation, it attracted much atten- tion from the loud and frequent peals of thunder accompanying its vivid lightning flashes, from its heavy rain, and in several places from the violent wind-squall from the northwest that accompanied it. A remarkable number of destructive lightning strokes appear on the records, and hail was reported from Lowell and Magnolia, Mass. The observer at Longmeadow, Mass., reported " low clouds flying up from the south" just before the rain began, while the body of the storm passed eastward to the north of him.

The rest of the 22d was showery and sultry ; several showers passed without thunder, but there were three distinct small thunder-storms among them. The first began somewhere east of the Connecticut Eiver, and moved eastward along the northern boundary of Massachu- setts over Newburyport between 7^ and 1 0'', with a velocity roughly estimated at 30 miles an hour. It was of short duration and covered a narrow path, but the times of rain beginning imply an oblique atti- tude of the rain-front, like that of the previous larger storm. Its rain was heavy, though brief, and its lightning struck in several places ;

* Two reports for the 20th are almost certainly misdated for the 19th, with whose storms they agree very well ; and one report, dated 1 a. m. Aug. 21, pretty surely belongs twenty-four hours later.

OP, ARTS AND SCIENCES. 37

notably at Wenham, Mass., five miles or more south of the rain area indicated by the records. The second storm began in central Massa- chusetts a little before noon, moved 30 to 35 miles an hour a little south of east, passing Boston at 12^30, Plymouth at IS''. 40, and Provincetown at li'^.lO; like the preceding storm, its rain was brief and heavy, with a well-marked northwest or west squall, and at Ply- mouth there was some hail at lo^Ao. The oblique position of the storm-axis with respect to its motion, as indicated by the times of rain beginning, is in this case directly confirmed by the record from Prov- incetown, where the storm-cloud was described as about three times as long as it was wide, with its longer axis standing west to east, while it moved to the southeast " with a sidewise drifting motion." The third shower was a very mild one, beginning a little west of northern Rhode Island about IS'^.SO, and crossing southeastern Massachusetts between 15'^ and 16\

August 23. No reports were received for this day; the distribu- tion of pressure gave gentle northward gradients, without any centre of low pressure being visible. /

August 24. On this day the gradients to the northeast were better defined, between an area of high pressure on the Southern States and a centre of low pressure that moved across the northern part of the Gulf of St. Lawrence. A rather remarkable series of storms occurred in the afternoon, as these conditions gave way before the approach of a tropical cyclone, whose centre lay off the coast of Georgia on the morning of the 2oth. The morning was fair ; soon after noon showers entered western Connecticut and Massachusetts, of moderate intensity at first, but gathering strength notably in two cases as they advanced eastward. In Connecticut, the separate showers were so numerous that they cannot be very satisfactorily traced. The first one traversed the State from northwest to southeast corner, at a rate of twenty miles or more an hour ; it had heavy rain, some hail, and a strong north- west squall at Norwich and on the border of Rhode Island. In this and some of the later showers there were many lightning strokes. In northern Massachusetts, a storm gathered east of the Connecticut Valley about 14'' and became very violent on the shore between Boston and Cape Ann about 16\30; the rain, wind-squall, and light- ning were all exceedingly strong as this storm ran off the coast to the east-southeast. Again, a storm that was described only as a shower in central Rhode Island about IG"", became very energetic about New Bedford at 18''. In the evening, the winds that had been southwesterly, bringing high temperatures (mean max. 82°), changed

38 PROCEEDINGS OF THE AMERICAN ACADEMY

to northeasterly, and later at night rain began, continuing into the next day (August 25) ; this was evidently the approaching rain area of the cyclone on the southern coast.

Auf^ust 25-30. The 25th was a cool day of northeast rain without thunder, the centre of low pressure being on the southern coast; and with the coming of this cool general storm the more continuous hot weather of summer was ended (mean max. from Aug. 25 to 30 not over 70°), and thunder-storms were practically over for the season. No thunder was reported during this period.

August 31. Eight scattering reports on this day, when a faint barometric depression with weak gradients stood in the north, close the records of the three summer months.

From the large mass of material collected, the records for July 9th and July 21st have been chosen for fuller discussion.

Storm of July 9th. — The large number of reports received for this date, and the violence of the storm described in many of them, jus- tify its special description. Its rain was generally heavy, though not great in amount, on account of its rather short duration ; its winds were often destructive, and at two places, Kent's Hill, Me., and West Brookfield, Mass., they developed into distinct tornadoes, with the funnel cloud and its dangerous accompaniments. We have here, as on all other dates, greatly to regret the absence of reports from northern New England ; they are so few that it is impossible to trace the storm across northern Vermont and New Hampshire with much confidence; but farther south its attitude is defined with considerable accuracy, as shown in Figure 4, Our earliest records come from Bur- lington and Charlotte, Vt., on the eastern shore of Lake Champlain, where the storm was felt about ]3'\ At 15'^ it stretched northeast- ward from southern Vermont to the angle of the Androscogtrin River in Maine, and at this hour there appeared to be a division of the storm by an east and west line a little north of Concord, N. H. ; the northern portion being somewhat in advance of the southern. The S. W. to N. E. attitude of the front was maintained with tolerable regularity, but the rate of advance seems to diminish as the storm fiided away after sunset on nearing the southeastern coast ; on the coast itself several stations report " no rain."

Although the front of the storm extended from southwest to north- east, there is good reason for not giving it a direction of advance at right angles to this line; many observers described its motion as from west to east ; and the crowding of the lines representing the succes- sive attitudes of the rain-front where they trend east-northeast, as in

OP ARTS AND SCIENCES.

39

Connecticut and Rhode Island, compared with the greater distance between them where they turn more to the north, suggests an advance oblique with the front of the stonn. Taking east by south as its direction of motion, we find an average velocity from southern Ver- mont to Cape Cod of about thirty-seven miles an hour ; but if the quarter-hour front-lines are to be trusted, this rate varied from over fifty in southern New Hampshire down to about twenty on Cape Ann, Mass.

Fig. 4.

The frequent mention of the storm's dividing as it approached an observer is not fully explained by the observations thus far gathered. It probably means that the intensity of the storm varied along the front line, and varied again as the storm advanced ; future observa- tions must decide this. The individuality of the storm was, on the whole, well preserved : certain features were observed with consider- able uniformity at a good number of widely separated stations, and if the observers had all been equally watchful, the record of this uni- formity would doubtless have been much increased. To illustrate this statement, the following extracts are made from the records, as far as possible in the observer's words.

Clouds before the Rain. — The high, forerunning cirro-stratus over- flow was observed well at several stations ; it is of especial interest

40 PROCEEDINGS OP THE AMERICAN ACADEMY

in connection with the theory of storms, and its first appearance on hot summer days, its gradual extension, and the motion of its fila- ments as compared with that of lower clouds, deserve careful watching. The rise of the cumuli, or " thunder-head " clouds, should be noted also, and their angular attitude and apparent direction of motion recorded several times.

Northfield, Mass.: Thinnest high clouds had passed zenith before 14.08; black bank of cumuli in N. W. at 14.50; rain about 15.40. Greenfield, Mass.: Clouds gathering rapidly at 14.00; rain at 15.50. Athol, Mass. : Clouds rising in N. and N. W. at 13.00; rain at 16 01 Deerfield Centre, N. H. : A little cloud in W. at 14.00; large black cloud in W. at 14.30 ; '■ clouds overhead," 14.55 ; rain at IQ.'IO. Man- chester, N. H. : Clouding up for three hours before rain began at 16.22. Tyngsborough, Mass.: Thin cirrus clouds rising iu S. W. and N. W. at 15.00 ; cumulo-stratus low in west at 15.50 ; the cirrus covers eight-tenths of sky at 16.10, and top of cumulo-stratus estimated 20° above horizon ; at 16.30, 60° over horizon ; at 16.37, edge overhead ; at 1 6-39 rain began.* Hartford, Ct. : Western sky hazy at 14.00, tem- perature 93°+, with brisk southerly wind ; a few light cumuli floated rapidly from S. W. to N. E., and a bank of cirro-stratus was seen low in the N. W. ; at 15.15 the bank had risen somewhat, aud cumuli were seen beneath it; at 16.30 cumuli could be seen rising from the cirrus; thunder was then heard and the sun was soon obscured. Clouds were rising in the west, but the storm seemed to be passing from S. W. to N. E. At 17.00 the storm seemed to approach the city, and at 17.23 the rain began and the wind veered from S. W. to N., blowing brisk ; later, the wind turned to E. ; plentiful large hail-stones and destructive wmd were reported a few miles east. Coventry, Ct. : Heavy clouds in N. W. at 15.30 ; rain at 17.30. Lynn, Mass. : " Edge of storm over- head" at 16.00 ; rain at 17.30. Chelsea, Mass. : Hazy clouds in W. at 12.30 ; heavy clouds in W. at 16.30; sky half covered at 17.10; rain at 17.32. Swampscott, Mass.: Heavy clouds rising in W. at 15.30 • heavy shower in W. at 16.00; destructive N. W. squall at 17 25; rain

* By means of these observations of altitude, it is possible to determine the height of tlie clouds, when the velocity of the storm's progress is known. In this case, the velocity was about forty-eight miles an hour for this part of the storm; and a simple construction then gives the heights as about eight and nine miles, at the times of the two observations. These, however, are not even in spite of their good agreement, sufficiently accurate or certain to base much argument upon; but they serve well to sliow how easily observations may be made that will lead to most interesting results.

OP ARTS AND SCIENCES. 41

at 17.35, in torrents at 17.40. Beverly Farms, Mass.: S. W. breeze and small black cloud in W. at 15.00 ; cloud grew till it extended along W. and N. horizon ; at 1 7.00 thunder-clouds moved rapidly towards S. E. ; at 17.10, edge of cloud in zenith, wind suddenly lulled, but a furious W. squall came up at 17.15 ; rain began at 17.35, and from 17.45 to 18.00 the rain poured in torrents. Quincy, Mass. : Clouds first seen in N. W. at 15.35 ; high N. W. wind at 17.32 ; rain at 17.49. Bar Harbor, Me.: Dark cloud in N. W. at 15.30; clouds darker and a heavy squall at IG.OO; heaviest wind at 17.30; rain began 17.50. N, Attleborough, Mass. : Thunder-caps just visible over N. W. horizon at 14.30; "edge of cloud overhead " at 16.55; rain began at 17.55. Taunton, Mass. : A stratum of hazy clouds in N. W. at 13.00; rising higher at 14.00; a lower stratum appears at 15.00; rain at 18.05. S. Weymouth, Mass. : Clouds rising in N. W. and thunder heard from them at 16.15 ; rain at 18.15. N. Marshfield, Mass.: Heavy clouds in N. W. at 16.45; rain at 18.15.

It appears from this summary that the cirrus overflow was often seen two or three hours before the rain began ; its edge passed the observer's zenith from an hour to an hour and a half ahead of the j-ain, and therefore probably extended thirty to fifty miles in advance of the main mass of the storm cumulus. The thunder-clouds them- selves were often visible an hour or more before the rain fell. The edge of the heavy cloud generally passed the observer's zenith three to twelve minutes before, that is, two to six miles ahead of, the rain- front; at several southeastern stations, where the storm's intensity (and velocity ?) had decreased, this interval was over half an hour. At one station (Oilman ton, N. H.) the rain began fifteen minutes before the edge of the cloud passed overhead ; and this was, as the observer well remarks, the more surprising, as no high wind occurred with the storm there.

The occurrence of " pocky " or " festoon " clouds (smooth masses hanging from the under surface of broad cirro-stratus, convex down- wards, like inverted cumulus clouds) is indicated by the following: " Rockland, Mass. : The setting sun was shining upon the under re- treating edge of the storm-cloud and showing all the irregularities of its usually smooth under surface, — a very pleasing sight." These clouds are of relatively rare occurrence ; they seem to be formed most commonly in connection with thunder-storms, and they should be carefully noted. The squall cloud is described in the account of the wind. '

42

PROCEEDINGS OF THE AMERICAN ACADEMY

TABLE II. Relation of Wind and Rain : Storm of July 9, 1885.

station.	Wind. 0-5.	Time.	Dura- tion.	Raia. Inches.	Time.	Amt. per hour.	Remarks.
Burlington, Vt.	S. 4	13.00	_	0.46	12.45-13.30	0.61	Much damage.
Bridgtou, Me.	-4	—	— .	0.37	15.00-17.00	0.18	(( (i
Chesterfield, N. H.	N.W. 4	15.20	20	S	15.18-15.55	14-	it ti
Brattleboro", Vt.	N.W. 3	15.12	20	0.45	15.21-15.55	0.77
Ac\7orth, N. II.	N.W. 4	15.00	45?	i Light.	15.30-16.20		Much damage.
Northfield, Ms.	N.W. 4+	15.35	10?	0.80	15.35?-16.16	1.18	Destructive.
Kent's Hill, Me.	W. 4	16.15	10?	i	15.40-16.25	i	Tornado near by.
Ashburuham, Ms.	N.W. 4	16.10	10	Mod.	16 13-17.05		Trees broken.
AUentown, N. II.	W. 4	16.39	8	1	16.43-17.20	§
Tyngsborough, Ms.	W.N.W. 4	16.37	6	Heavy.	16.43
Lowell, Ms.	N.N.W. 3	16.45	15?	—	16.49-18 00
Westford, Ms.	W.N.W. 4	16.43	15?	i	16.43 18.15	*
Millville, Ms.	N.W. 4-1-	17.20	20?	A	17.35-18.15	13	Almost a tornado.
Beverly Farms, Ms.	W. 4-f-	17.15	30?	i	17.35-18.00	¥	Furious squall.
Bar Harbor, Me.	N.W. 4	17.30	—	Light.	17.50-18.20		Very heavy squalls.
Providence, R. I.	N.W. 4	17.55	13	0.57	18.08-19.10	0.55	Almost a hurricane.
Rockland, Ms.	N.W. 4	17.43	15	Light.	(18.00-18 05 f / 18.21-19.05 i

The most obvious relation of wind and rain is, that a brief heavy squall shortly precedes a heavy rain-fall ; and this is so commonly stated to be the case that one is tempted to regard the first as the effect of the second. But the relation has two exceptions : first, as in the above table, where heavy winds occur with light rain, although in those cases heavy rain may have occurred near by ; second, in other storms there is often heavy rain and no severe wind. More data are needed before attempting further generalization.

It has been stated that a ragged fringe of cloud always advances in the front of a storm that is accomj^anied by a squall.* The observa- tions made at a number of stations seem to confirm this opinion. The following may be quoted : Northfield, Mass. : A bank of inky clouds rose from the N. W. bordered by a line of fog ; a violent gale sprang up suddenly ; soon after this, the line of fog passed the zenith, and the air was suddenly darkened. Amherst, Mass. : A scud of low black clouds was blown rapidly from N.W. at about the time of high N.TV. wind ; heavy rain soon after. AUenstown, N. H. : An arch of boiling scud before the heavy storm-clouds, followed shortly by violent W. wind, with rain whirling in sheets. Tyngsborough, Mass. : Air calm ; high cumuli, "preceded by a ragged and turbulent gray squall-cloud"; seven minutes later, a heavy squall, air filled with dust, and " white caps "

- * Hann, Beitrag zur Morphologie der Gewitterwolken. Zeitschrift fiir Me- teorologie, 1880, xv. 434.

OF ARTS AND SCIENCES.

43

on Merrimack River. Ward Hill, Mass. : Black clouds rising, " with a swirling mass of light, smoky clouds along the whole length of the upper edge ; at 16.45, the bank was overhead, stretching from N. E. to S.W., and rushing on with great speed " ; about 17.00, a high S.W. wind. Lowell, Mass. : Dark clouds rising in N.W. ; at 16.44, " a white, foamy, frothy cloud, considerably disturbed, seemed to be pushed or rolled rapidly along ahead of the black clouds, and came towards us fast from the N.W. and N. " ; as it passed overhead, the wind changed from S. to N. with sudden gusts at 16.45, and heavy rain at 16.49 ; a relatively clear space was noticed between the squall-cloud and the heavy black clouds that followed rapidly behind and above it. This observer describes the squall-cloud as rolling along, the upper part seeming to pass forward over the lower. Special attention is desired on this point. Haverhill, Mass.: at 16.50, an immense windrow of white clouds, extending from the W. to the N. horizon, came literally rolling along toward our city ; behind it, a great area of black clouds with continuous lightning ; very high, sudden wind, with moderate rain, at 17.00.

TABLE III.

Change of Temperature in the Storm of July 9, 1885.

Station.	Before Rain.	Rain.	After Rain.	Change.
		h.	h.	h.	o m.
Pittsfield, Mass.	91 at 15.00	15.10	70 at 15.45	21 in 45
Athol,		82 " 16.01	16.01	67 " 16.20	15 " 19
Belcliertown,		90 " 16.00	10.15	72 " 16.30	18 " 30
it		95 " 15.00	16.15	70 " 17.00	25 " 120
Fitchburg,		92 " 1(5.08	16.16	74 " 16.22	18 " 14
Gilbertville,		90 " 16.00	16.28	68 " 16.50	22 " 50
i<		85 " 10.24	16.28	68 " 16.50	17 " 26
Monson,		88 " 16.45	16.50	72 " 17.05	16 " 20
Worcester,		88 " 16.49	16.51	73 " 17.02	15 " 13
a		94 " 16.30	16.53	71 " 17.15	23 " 45
N. Grafton,		88 " 16.53	16.58	74 " 17.06	14 " 13
Southbridge,		85 " 17.00	17.05	68 " 17.15	17 " 15
Newton,		81 " 17.18	17.18	72 " 17.30	9 " 12
Wate'rtown,		90 " 1645	17.20	72 " 17.30	18 " 45

A rapid fall of temperature is thus shown to accompany the storm- front ; its average rate is probably a degree in one to three minutes, but the most rapid change is much faster ; the total value of the change ranges from ten to twenty or more degrees. The cause of the change is to be found in part in the protection from hot sunshine by the storm- clouds, but it is probably much more dependent on the low temper-

44

PROCEEDINGS OP THE AMERICAN ACADEMY

ature of the cold rain and hail- that have been condensed high up in the clouds and chill the air as they fall through it. The relation of amount and rate of rain-fall, temperature of rain (or hail), and change of air temperature, form an interesting subject for study.

We have two valuable automatic traces of the temperature curve in this storm from the Boston Water Works at Chestnut Hill, Mass., and the City Engineer's office at Providence, R. I. The first curve shows a rapid fall from 86° at 17.30 to 69° at 18.00, followed by a tem- perature at midnight a little warmer than that of the early evening. The second gives a fall from 86° to 74° during a i'ew minutes before and after 18.00.

Barometric Changes in Storm of July 9. — The self-recording barom- eters (Draper's pattern) at Providence (City Engineer's office) and at the Blue Hill Meteorological Observatory were both affected by this storm in a similar way. The hundredths of an inch over twenty- nine inches are given in the following table for fifteen-minute intervals before and after the rain began.

Tl	^BLE IV.
Time.	Blue HiU.	Providence.
m.	in.	in.
120 before rain.	.085	.715
105 "	.085	.715
90 "	.090	.710
75 "	.085	.700
60 "	.080	.710
45 "	.065	.715
30 "	.080	.720
15 "	.100	.785
Rain began.	.115	.800
15 after rain.	.130	.800
30 "	.170	.825
45 "	.155	.775
60 " "	.140	.745

From this it may be said that there was a slight fall of the barometer before the storm, followed by a rapid rise as the storm came on, and succeeded by a fall again as it passed away.

A composite portrait of this storm has been prepared similar to the one here figured in illustration of the squall of July 21 (Fig. 6) ; its size is too large to admit of reproduction. By means of a simple process of construction, all the observations of wind, temperature, rain, thunder, etc., are thrown in their proper relative positions with respect to the

OF ARTS AND SCIENCES. 45

storm-front, so that a graphic average is obtained of the distribution of these elements in the whole storm. An hour and a half or an hour before the storm, clouds are seen rising on the western horizon, while the winds are light southerly, and the temperatures high (85-95°). Nearer the storm-front, the clouds are seen to rise higher, and the tem- perature falls slightly, but the wind does not change significantly. The first thunder is heard from thirty to sixty minutes, and the clouds are • recorded as reaching the zenith or passing overhead from ten to thirty minutes, before the rain. The sudden change from gentle southerly wind to the northwest squall seldom comes more than fifteen minutes before the rain, and is generally only five to seven minutes before it ; with this change the temperature falls rapidly. The squall seldom continues after the rain begins. The heaviest rain is marked close to the rain beginning in many cases ; in others, it fsills from seven to twenty-five minutes later. The loudest thunder runs from ten to thirty-five minutes after the rain-front, and the lightning strokes, as far as reported, fall with one exception between thirteen and twenty-seven minutes after the rain-front. Already at twenty to twenty-five minutes after the rain had begun, the western horizon is seen lighting up, and soon the clouds begin breaking away ; their rear edge is overhead iji an hour to an hour and a half, while the rain had ceased fifteen minutes sooner on the average, its shortest duration being thirty, and longest ninety minutes. During the rain, the temperature stood fifteen to twenty-five degrees lower than before the storm, and the winds were light and variable ; as the storm passed over in the afternoon, an abso- lute rise of temperature after its passage is seldom seen, and then is faint ; but a relative rise is clearly found in the maintenance of an almost uniform temperature past those hours when it ordinarily de- creases most rapidly. Rainbows make their appearance between an hour and an hour and a half after the rain beginning, and the last thunder is heard from one to two hours after the storm began. We gain easily, from such a portrait as this, a general view of the storm that can be acquired only with much labor in other ways. As to the accuracy of the view, that can be greatly increased as our observers come to use more uniform methods in making their records.

Storms of July 2\. — The thunder-squall that traversed southern New England on the morning of July 21 was in some respects the most interesting storm of the summer. The persistent maintenance of its several features throughout its whole path is especially instruct- ive. It closely resembles squalls that have been described in Iowa by Hinrichs.

46

PROCEEDINGS OF THE AMERICAN ACADEMY

The squall entered New England at the western end of the Massa- chusetts-Connecticut boundary about 10.10 a.m. (See Fig. 5.) Before this it had been noted at Hensonville, Greene Co., N. Y., where the rain began at 9.22, with a strong N.W. gust of wind, in which the temiDcrature fell ten degrees (81° to 71°) in two minutes; the storm seemed to pass centrally over the station, and its clouds were very black and angry. Still earlier observations were made in New York, as follows : Peterborough, Madison Co., loudest thunder at 7.22 in the south, with light rain at 7.45 ; Palermo, Oswego Co., storm to south with loudest thunder at 6.20 ; Constantia, Oswego Co., thunder first heard at 6.45, continuing for two hours. The last two reports

MEW/YORK /

I ' ■■■'■■/ I '. . w > *s • ^; .■^-- 'v ■ 1 J^t ■■-■■.'•■. ■"■»■•

CONNEA^TICUl

mt;

Fig. 5.

are furnished by Prof. H. A. Hazen, U. S. Signal Service, and are here mentioned with the rest to point out the probability that this storm either began at a very unusual morning hour, or else was a sur- vival from some storm of the previous afternoon. That they refer to the squall here described admits of no question, as they correspond closely to the position that it would have at their several times, as de- duced by backward projection from the New England observations. Returning now to our own records, we have Great Barrington, Mass., rain at 10.40 with heavy clouds and continuous thunder. West Norfolk, Ct., It. rain 10.30. Norfolk, Ct., rain by three observers 10.35, 10.40, 10.40, heavier to N. with high N.W. wind at 10.40. Winsted, Ct, 10.33, clouds rising hi N. ; It. rain at 11.08, with moderate N.W. wind; shower nearly out of sight at 11.35. Collinsville, Ct., black thunder- cloud to N. at 11.00, wind moderate S.W. shifting to brisk N. at 11.15, It. rain at 11.25. Longmeadow, Mass., dark clouds rising in W. at

OF ARTS AND SCIENCES. 47

10.43, soon rising very fast, bringing a strong W, N.W. gust with clouds of dust and leaves at 11.10; considerable damage to hay, grain, and trees was done by this wind ; rain began at 11.12, and about | inch fell in 18 minutes ; the storm was nearly out of sight in the E. at 12.00. Springfield, Mass., a heavy wind (no direction given) from 11.20 to 11.35 with a few drops of rain. Chicopee, Mass., It. rain with W. wind about 11.10. Broad Brook, Ct., violent wind and dust storm just be- fore noon. Monson, Mass., It. rain at 11.30, with brisk S. W. wind at 11.35. Tolland, Ct., black clouds came up very fast, It. rain and brisk W. wind at 1 1 .35, more rain to N. Quarry ville, Ct., rain at 11.45 with strong N.W. gust about noon. Southbridge, Mass., It. rain at 11.45 with fitful gusts from S.W. Dudley, Mass., It. rain at 11.55 with S.W. wind. Thompson, Ct., a shower came up very quickly from a little N. of W. about 11.45, violent but short-lived; rain-fall mod- erate ; lightning struck a church spire. Danielsonville, Ct., It. rain (no time given), but heavier to N., with thunder from 11.30 to 13.00. S. Killingly, Ct., rain with sudden heavy gust at 12.10.

The storm now enters Rhode Island. Pascoag, rain at 11.55, and high wind from W. N.W. ; at 12.40, thunder-clouds receding in the east. N. Scituate, clouds passed the zenith at 12.15, It. W. wind; sudden rain, heavy thunder and high W. wind at 12.19; rain over at 12.45. Woonsocket, rain and brisk S. wind at 12.20; rain in torrents at 12.22. Ashton, clouds rising rapidly in W., temp. 96'^; rain at 12.25, and brisk W. wind at 12.30. Pawtucket, very violent N.W. squall and heavy rain at 12.27, temp. 94° ; at 12.45, temp. 77° ; at 12.58, rain over; It. "W. wind, temp. 84°, and dark clouds in E. and S.E. : second observer ; very dark wild-looking clouds, moving rapidly from W. to E., rain in torrents with very high W. wind at 12.29il ; at 13.00, sun shining hot. Providence, five observers : 1st, heavy wind-clouds in N. W. at 12.19 ; clouds passed zenith and moderate rain began with high N.W. wind at 12.27 ; clouds disappeared by 13.15 : 2d, heavy clouds in W. and N.W. at 12.08 ; clouds at zenith at 12.13 ; high west wind at 12.26, rain at 12.28 : 3d,' temp. 93°, with moderate S.W. wind at 12.15, It. rain at 12.28, high W. N.W. wind at 12.30, temp. 88° ; at 12.45 temp. 77'=: 4th, at 12.20 heavy gusty wind; brisk rain at 12.30: 5th, cumulo-stratus cloud rapidly forming in W.> at 11.00 ; fresh W. S.W. wind, temp. 90" at 12.00 ; very strong dust squall at 12.27, and heavy rain at 12.30 with edge of storm overhead ; temp. 74|° at 12.37. Sil- ver Spring, edge of cloud overhead, with high W. wind at 12.30 ; very heavy W. N.W. squall at 12.34; moderate rain at 12.35 ; blue sky in W. at 12.49 ; rain ended, dark clouds in S. E. at 12.58.

48 PROCEEDINGS OF THE AMERICAN ACADEMY

The storm next enters southeastern Massachusetts, and at the same time extends northeastward suddenly. N. Attleboro', shower came up very quickly ; at 12.00, temp. 98° ; edge of cloud overhead at 12.20 ; It. rain and very liigh W. gusts at 12.30 ; temp. 84°, 12.45. Attleboro', storm at its height with W. wind about 12.25. Norton, temp, before the storm 94°, the hottest of the season; about 11.00 clouds arose, wind S. ; about 12.00 a dense black "fog" rose in the west ; rain at 12.15; wind increased, with terrific lightning and thunder; at 12.20 the " gale " struck in full force, lasting 5 to 8 minutes, breaking down and uprooting trees ; the track of the gale was about I mile wide, ex- tending from W. N.W. to E.S.E. Mansfield, smart shower about noon, house struck by lightning. Bridgewater, black clouds seen about 1 2.00 ; hurricane of dust and pouring rain about 12.45. N. Easton, heavy shower and destructive gale about noon. Raynham, rain at 12.15 ; high wind and hard rain at 12.30. Taunton, 1st, clouds rising in W. 12.20 ; moderate rain and high W. wind, 12.40; light in west, 12.50: 2d, storm came up quickly from N.W. ; heavy rain and destructive S.W. wind, 12.50 : 3d, It. rain with very high W. wmd at 12.55. Middle- borough, damage by wind and lightning. Lakeville, brief destructive storm. Fall River, heavy rain in W. at 12.35, temp. 87° ; high S.W. wind with a rush " like a squall contained in a large black cloud " at 12.55 ; It. rain began at 12.57 ; blue sky in W., temp. 77°, at 13.15. Pembroke, rain and brisk S.W. wind at 12.42, temp. 92° ; hard rain, 12.52; rain ceased, temp. 81°, 13.00; heavy black clouds in the S. Plymouth, high S.W. wind at 12.45 ; hard rain at 12.50 ; violent S.W. squall at 13.00 ; damage by wind and lightning. E. Freetown, a brief shower with light rain beginning at 13.04, came up very quickly. Long Plain, tree struck by lightning. New Bedford, shower heavier to N. ; It. rain at 13.15. Cotui't, brief shower at about 13.30. Oster- ville, heavy clouds in N. at 13.00 ; It. rain with no wind at 13.50 ; clear in W. at 14.00. Provincetown, It. rain at 13.31 ; high W. S.W, wind at 13.37; clouds moved from W.N.W. ; clearing at 14.00. Adjacent stations reporting " no rain " or " distant thunder " are shown in Figure 5.

These extracts are given as nearly as possible in the words of the observers ; they embody but a small part of the observations recorded, but represent all the stations from which definite statements have been received. They are given here, not only to illustrate the character of the reports that have been made, but also to show in detail, for one storm at least, the focts on which the generalizations as to form, path, and velocity are based. The importance of accurate time records is

OP ARTS AND SCIENCES. 49

especially evident in a fast moving storm like this one, where the ve- locity was almost a mile in a minute : an inaccuracy of ten, or even of five minutes, would cause a considerable error in drawing the storm- front curve. The incompleteness of records is especially to be regret- ted; for example, at Springfield two observers failed to make any report, and the brief statement for that station is taken from the regu- lar record of the observer of rain-fall, who reports for our monthly Bul- letin. Hartford makes no mention of the storm, although it must have passed within sight, if not within ear-shot. A number of stations in central Rhode Island were also silent. Besides this, the occurrence of considerable stretches of country from which we have no records is especially unfortunate. Although the number of reports is consider- able, there are certain districts where they are lamentably insufficient, and, in constructing the maps to illustrate this storm, one has contin- ually to regret the lack of observers in numerous good-sized towns situated directly in the path of the squall.

Figure 5 represents the path and progress of this storm, as deter- mined by charting all the observations collected. Stations where rain was noted are marked by a black dot. The curved dotted lines show the supposed attitude of the rain-front for every quarter-hour. Arrows indicate the velocity and direction of wind accompanying the rain. A number of valuable negative records, such as "no rain " or " distant thunder " while the storm was passing, are represented by small circles, with T on the side from which thunder was heard. As thus deter- mined, the average velocity of the storm's progress is forty-eight miles an hour ; its path lay a little south of east ; it seems to have run faster in eastern Connecticut, and on entering southeastern Massachusetts it suddenly extended to the northeast. Besides the larger storm of the afternoon, three others were observed on the morning of this day ; one in southeastern New Hampshire and northeastern Massachusetts ; another to the southeast of Boston, which may have united with the principal squall out at sea ; but while on land they were distinct, as is proved by the three " no rain " stations between them ; the third was a small storm in southern Rhode Island, about 13.00 to 14.00 hours. Return- ing again to the chief storm of the morning, we find that in south- eastern Massachusetts the scanty time records do not suffice to define the storm-front, which there seems to become irregular ; but elsewhere the storm was strongly convex to the east, with its most violent action in rain, wind, and lightning at the apex of the curve. All observers agree that the storm came and passed away quickly ; the rain-fall sel- dom measured over a quarter of an inch, but the fall was often heavy

VOL. XXII. (n. S. XIV.) 4

50 PROCEEDINGS OP THE AMERICAN ACADEMY

in its brief duration. The radially outblowing wind, of destructive strength, at and a little south of the apex of the storm-curve, was aa invariable accompaniment of the storm for at least two hundred miles aloD"- its path. It was doubtless felt before the storm reached Henson- ville, and after it crossed Massachusetts Bay and ran out to sea beyond Cape Cod ; it would be interesting to obtain a record of it in these at present unknown parts of its path, to see how nearly they would fall in with the observations here charted. A rapid fall of tempera- ture accompanied the fall of the rain, as will be further shown below. The Catskill Mountains, over 2,500 feet high, the Hudson valley, the mountains of western Connecticut and Massachusetts, with summits up to 2,500 feet, and the Connecticut valley, were all traversed by the storm, without exerting any definite effect on its course or velocity. Whatever the mechanism of the storm may be, it is evident that it is at work chiefly at a considerable altitude, and that it is borne along by the general winds in which it is engendered, while only its more remote effects reach the ground.

A " composite portrait " of the surface effects of this storm is pre- sented in Figure 6. It is constructed by throwing all the observations (except certain ones in S. E. Mass.) into their proper place with respect to the rain-front and middle-path.* The spaces between the curved lines represent intervals of fifteen minutes before or after the time of rain-beginning, and hence correspond to a distance of about twelve miles, as the velocity of the storm was forty-eight miles an hour. All observations of a single station fall on a line to one side of and parallel to the middle-path, and at their appropriate interval before or after the rain-front. The temperatures are given by num- bers, showing degrees Fahrenheit. The winds are marked by arrows whose feathers increase with their velocity; arrows without bai'bs represent winds when force but not direction was given. The first, loudest, and last thunder are indicated by Tj, T, and Tg. A few lightning strokes are marked by L. The condition of the sky is roughly shown by black or white crescents, which mean clouds or clear sky in the direction of their convex curve. The dui-ation of the rain is marked by black lines. The portrait as thus constructed there- fore gives a graphic generalization of the average relations among all these features of the storm.f In this case the material is insufficient for a wholly satisfactory diagram ; but a good purpose is served in

* See Proceedings of the American Academy, 1885-86, xxi. 346. t Figures 5 and 6 are reprints from Science, May 14, 1886.

OP AETS AND SCIENCES.

51

a2

^

1 ;m

',";"

I i

D. >- ^ If ^

O ^ 3 tt.

t ' i 1,1! . t i2 c

,T' .'(1-. ^♦v-'g

s.

'•--'lit' '^1 -> ' I ' • I fy i. VIA

I »

Fig. 6.

52 PROCEEDINGS OP THE AMERICAN ACADEMY

bringing all the data together in a convenient shape for further exami> nation. The errors of the method consist in the assumption of uni- form shape of the storm-front, and a uniform velocity of progression; but I do not think they are serious in this case. The convexity of ^he storm-front is perhaps too great, but that it was really well bowed forward is shown by the late arrival of the rain at the lateral, as com- pared with the central stations. The concavity of the rear of the rain area is not so well established, but it seems to have approached a parallelism to the front.

83'

^Z' -80-' "~-.^86'/a" ,^0' "-vSr

- -^-H ''- - .\-7 -"^--V-4

NO »- ^ ^-*| '

y

-- X ..>..\.6 _-- ■^-z .--... ^.\a.~.- ...^.Vl-

\ -s- - *- - -V<i ■ • r ■ - - --«'< • • •"'Sacr-.- -\'' -

'. * » * ' . ;-a ..,-s ^„ ,...,. ..^o --V,

t /:^:::;-.\;^^^r^ ■^^-^.-- ^^^^-^-i^-^s c-.-:^::^?

/ ' ''- ^ ♦ / /

/

,.i, . y.,3 ,.. -A — y^,s _ . . ^ ^f/7 ..-.-..-.-. -.,-^o .J:vr.-.;^o°. ■--

„^^ --ji-- 7-^-16 -is---i^^-t -^o .,•'♦( — :_

--->'-6 .....^<J .„ ^- #-*- 1-— ». --^o — z,'*}^-"

Fig. 7.

The changes of temperature with the passage of the storm are well determined. For their better illustration I have prepared Fig. 7, as follows : by interpolation, an approximate temperature was determined for a number of stations at the several fifteen-minute lines of Fig. 6 ; then, taking the temperatures at fifteen ramutes before the rain as standards, the departure from them was determined for the other times, and these plus and minus numbers are exhibited in Fig. 7. Averaging the departures and the observed temperatures on every time line gives the following results : —

Interval.	Departures.	Mean Temperatures
30 min. before rain. 16 "	+r	91° 90
Rain begins. 15 min. after rain, 80 " 45 " "	— 8 — 7	86i 80 82 83

OF ARTS AND SCIENCES. 63

There is some indication that the greater changes of temperature are near the apex of the curve, as would be expected. We are for- tunate in having two records close to the apex that show the greater part of the temperature fall to have been very sudden ; one is at Hensonville, N. Y., as follows : —

Time 9.13 9.20 9.25 9.27 9.30 9.35 9.40 9.50 10.8 13.30

Temperature 85° 82° 81° 71° 70° 71° 72° 73° 74° 92°

Wind — — Strong. — — — Light. V. light. — —

The second is from the City Engineer's office, Providence, R. I., where a self-recording thermometer (pattern of Richard Freres) gives a curve showing a fall of 13° in thirty minutes, the greater part of the change being accomplished in a small fraction of this time. A second fall at 1G.35 was caused by the large afternoon storm of this date. From the same source we have the automatic record of a Dra- per's mercurial barometer, showing the abrupt increase of pressure characteristic of thunder-storms, from 12.33 to 12.40 the pressure rising from 29.695 to 29.730; and also an automatic wind record, showing the sudden wind-squall at Hope Reservoir, Providence, blowing at a rate of three miles in the five minutes from 12.33 to 12.38; before and after the squall, the velocity was ten to twelve miles an hour. Another barometric record, from a Richard Freres aneroid, was obtained from Plymouth, exhibiting an abrupt rise of one millimeter = 0.04 inch at a few minutes after 13.00. It is greatly to be hoped that self-registering instruments — especially barometers — may be more generally kept at our stations: considering the great value of their records, their cost is not relatively high. Detailed observations of the clouds of this storm were so few that no generali- zations can be made from them.

The afternoon storm of July 21 (Fig. 8) came from New York, crossing the Hudson mostly below Albany. The temperature in New England had risen over 90°, with southwesterly winds, in many places after the passage of the noon squall ; the heat became very oppressive before the larger storm arrived. The thin front of the high cirro-stratus was visible in the northwest at least three hours before the rain began, and as it obscured the sun the temperature fell slowly. The front edge of the lower clouds stretched S.W. to N.E. in Connecticut, where the short-lived squall-wind attained a destruct- ive strength, and was accompanied with hail at a number of points ; but, unfortunately, our observers there were few, and many records had to be gathered simply from the current reports of the newspapers.

54

PROCEEDINGS OF THE AMERICAN ACADEMY

New Britain, Willimantic, and New London all had a violent storm, and a belt of destructive hail fell between Middletown and Colchester. In Massachusetts the storm was brief, but rather violent around Worcester, where numerous lightning strokes were reported ; but the distribution of these strokes, as well as of the hail, is not based on sufficiently uniform reports to serve as the ground of any general statements. On Blue Hill the wind was very violent for a short time, but elsewhere in eastern Massachusetts it was generally of moderate severity. The backward turn of the rain-front in Massachusetts began too far inland to be attributed to the effect of the ocean ; it is rather to be regarded as the ordinary lagging behind of the side of

fy^^i

.«

.^\

IB

li

\; j..e-' i).i-i --;•;' 4 : .--''4' ---' / • /i/ - d

'■'■■'&.■■■"

19

Fig. 8.

the storm. The wind-squall at a number of the marginal stations turned from its general northwest direction, and blew out from the side of the storm to the northeast or north. The observations off shore, on Block Island, Martha's Vineyard, and Nantucket, and at Provincetown, accord tolerably well with the advance of the storm as determined on land ; and Provincetown reported seeing lightning in the east till late in the evening.

The storm seems to have advanced to the east-southeast at an average rate of forty-three miles an hour. Like the storm of the morning it is peculiar in approaching the sea-coast with hardly diminished severity.

Revieio. — In view of the many questions that still remain unsettled, it does not seem safe to enter yet very far on the interesting task of

OP ARTS AND SCIENCES.

65

generalizing the results obtained ; but a brief statement may be allowed in review of the foregoing pages.

Table V. gives the hours of " loudest thunder " for June, July, and August, 1885, and their percentages; also the totals for the three months. Owing to the growth in the number of observers through June, and the omission of a number of reports not adapted to being counted here, the table cannot be taken as giving a precise view of the relative frequency of storms in the three summer months ; and, again, on account of the smaller number of observers who record the storms that occur at inconvenient hours, the percentages of night observations are probably below their true figures. Nevertheless, the general dis- tribution of storm occurrence is clearly made out.

TABLE V. — Hours of "Loudest Thunder."

	June.	July.	August.	Total. 1
Hours.
Obs.	%	Obs. ] %	Obs.	%	Obs.	%
A. M. 0-1	1	0.18	3	0.35	3	0.51	7	0.35
1-2	0	0.00	3	0.35	10	1.70	13	0.65
2-3	5	0.91	3	0.35	8	1.36	16	0.80
3-4	11	2.00	5	0.58	9	1.50	25	1.25
4-5	27	4.90	5	0.58	10	1.70	42	2.10
5-6	47	8.64	3	0.35	16	2.72	66	3. .30
6-7	1	0.18	4	0.47	5	0.85	10	0.50
7-8	4	0.73	5	0.58	12	2.04	21	1.05
8-9	8	1.45	12	1.39	27	4.50	47	2.34
9-10	18	3.27	5	0.58	18	3.00	41	2.05
10-11	7	1.27	11	1.28	13	2.21	31	1.55
11-12	8	1.45	26	3.02	3	0.51	37	1.85
P.M. 12-13	16	2.90	69	8.12	16	2.72	101	5.04
13-14	49	8.88	57	6.61	22	3.74	128	6.39
14-15	129	23.41	68	7.89	24	4.08	221	11.04
15-lG	106	19.06	130	15.08	38	6.46	274	13.69
16-17	88	6.90	164	19.02	79	13.43	281	14.04
17-18	56	10.34	153	17.75	101	17.17	310	15.48
18-19	8	1.45	78	9.05	79	13.43	165	8.24
19-20	2	0.36	37	4.29	47	7.92	86	4.30
20-21	3	0.54	14	1.62	33	561	50	2.50
21-22	1	0.18	6	0.70	12	2.04	19	0.95
22-23	5	0.91	0	0.00	3	0.51	8	0.40
23-24	1	0.18	1	0.12	1	0.17	3	0.15
	551	100.10	862	100.13	589	99.88	2,002	100.01

The excess in the later afternoon hours appears distinctly enough. But the time of greatest frequency does not depend simply on the hour of the day. If we take, on the one hand, southern Vermont, southwest- ern New Hampshire, western Massachusetts, and western Connecticut

66 PROCEEDINGS OF THE AMERICAN ACADEMY

as a western group of areas, and, on the other hand, eastern and south- eastern Massachusetts and Rhode Island,* as an eastern group, the hours of maximum- percentages come out as follows: —

Area.	June	July,	August.	Mean.
Western,	13-15	15-16	16-17	15-16
Eastern,	14-16	18-19	18-20	18-19

This clearly implies that distance from some at present unknown starting-point exerts a decided influence on the hour of maximum storm frequency ; or, in other words, that most of our New England storms come to us ready made from a source where they begin at a rather uniform hour of the day. Another season's work may perhaps determine where our storms begin. The importance of this question in the attempt to predict thunder-storms is evident. In Bavaria, Von Bezold finds a second faint maximum in the early morning hours, and attributes it to the arrival of storms from a source more distant than that which furnishes the more numerous ones of the afternoon. Our faint secondary maximum, at five to six o'clock in the morning, may have a similar meaning ; but the question cannot l)e decided from the observations of a single season, when one or two storms would exert a strong effect on the total percentage.

A brief examination of the " general account " suflfices to show that, while the summer storms there recorded are much more frequent than they are in the winter months, still they are not at all uniformly distrib- uted through the summer season : they appear in greater number and size for a few days, and then are almost or quite absent for a time. The cause of this seems to be found in their dependence on the larger dis- turbances in the atmospheric circulation, for which I should wish to use the name of " cyclonic storms" (following Piddington, Redfield, Ferrel, and many European writers), but which are perhaps more generally known by the phrase " areas of low pressure," as used in the Signal Service publications, or as " barometric minima" in Germany. It is as a rule only when we stand in a certain attitude with respect to the cen- tre of these low pressure areas that large and well-developed thunder- storms appear. It may therefore be concluded that the development of such storms depends, not only on the heat of the summer afternoons, but also on the equilibrium of the atmosphere as determined by the

* Fourteen reports of the squall of July 21, falling between 12'i and 1S\ are omitted in this count.

OP ARTS AND SCIENCES.

67

circulation of cyclonic winds.* The greatest number of our summer thunder-storms occur when a centre of low pressure lies to the north or northeast of us. But there are also examples of storms occurring in areas of high or equable pressure, as on August 1st; these would seem to correspond to the Wdrme-Gewitter of Mohn, and to depend on a local warming of the lower air sufficient to bring about the un- stable equilibrium that in the other cases ( Wirbel - Gewitter) depends on the arrangement of upper and lower currents m the cyclonic circula- tion. A more peculiar exception to the general rule of occurrence is found on certain dates when a well-developed low pressure centre passed north of New England, without bringing any local storms to us. The most pronounced example of this condition was on July 17th, in the middle of a long period remarkably free fi-om storms ; it was one of the hottest days of the summer, and yet passed without reports worth mentioning. Other exceptions have a more visible explanation: on June loth, low pressure centres stood both northeast and northwest of us, and no storms occurred ; on July 10th, low pressure centres were north and south of us, and we had only light rains ; the occurrence of two neighboring low pressure areas may in these cases have interfered with the arrangement of upper winds needed for the development of the heavy clouds of thunder-storms.

The direction of motion of thunder-storms seems to be about at right angles to the radius joining them with the low pressure centre ; that is, about parallel with the higher winds in which they are borne. This is clearly shown in the following table, embracing thirty-five storms whose course was sufficiently well determined to be used in evidence.

TABLE VI.

Motion of Thunder- storms.	Direction to Low Pressure Centre. 1
West of North.	North	N. N. E	N E.
North of East . . East South of East , .	7 1 0	1 4 2	1 2 7	0 4 6

* Tlie first definite announcement of tlie fact tliat thunder-storms have a definite position in areas of low pressure was made by Marie'-Davy, of the Frencii meteorological service, in 1864 : " The appearance of storms always coincides with the presence of rotary winds known under the name of * bour-

#

58 PROCEEDINGS OP THE AMERICAN ACADEMY

The relation of the velocity of local storms to the gradients and the velocity of the cyclonic areas in which they occur, needs still further investigation. The same may be said of the conditions that determine the arrival of some storms early in the day, and of other storms late; the passage of some storms with apparently undiminished strength off the sea-coast and their endurance into the evening, while others act in the opposite way ; and the maintenance of high temperatures after some storms (morning squall of July 21, and 'morning storm of August 4), while most are followed by decidedly cooler weather.

The surface winds flowing towards the storm at a little distance from it, and the squall commonly met blowing outwards in front of the rain area, are well defined on many occasions. The relation of the rain to the out-blowing squall is variable ; in many cases, the former seems to be the effect of the latter ; but in others we have heavy rain and no squall ; sometimes the squall is felt at only a few stations, while the heavy rain is general. Many observers speak of a connection between lightning flashes and an increase in the rain-fall immediately following; it is quite possible that this relation exists, but the evidence usually quoted to prove it does not seem conclusive.

The more general distribution of storms in the areas of low pressure, their total duration, and the possibility of their surviving the cooler hours of night, are problems that can be better determined by the work of the Signal Service in reviewing the more detailed studies of local services. It is hoped that many of these problems may find their solution in that larger fund of material towards which this report is offered as a contribution.

rasques.' " This conclusion has been confirmed by other European services, and by Prof. H. A. Hazen in this country. Lieut. Finley has shown that it obtains also for our tornadoes.

Cambkidge, Mass., June, 1886.

OF ARTS AND SCIENCES. 59

Til.

CONTRIBUTION FROM THE HERBARIUM OF HARVARD

UNIVERSITY.

A PRELIMINARY SYNOPSIS OF NORTH AMERICAN

CARICES,

INCLUDING THOSE OF MEXICO, CENTRAL AMERICA, AND

GREENLAND, WITH THE AMERICAN BIBLIOGRAPHY

OF THE GENUS.

By L. H. Bailey, Jr.

Presented April 14, 1886.

In the following Synopsis I have divided the genus, or rather its American representatives, into two sub-genera and fourteen sections. The primary divisions of the sections, designated by capitals, may be called groups. The names of the groups are the plurals of specific names. Sectional names of various ranks are now so numerous, that I have made the endeavor to choose in accordance with recognized rules of priority. If I have seen and examined critical or historic speci- mens, the collector's name has been printed in Italics. In no case have I admitted uncertain authorities for geographical distributions. Herb. is an abbreviation for Herbarium. Distinguishing characters have been given for those species which are not described in Gray's Manual, Chapman's Flora, or Coulter's Manual of Rocky Mountain Botany.

A genus so vast as Carex must always suffer divisions which are founded upon appearances rather than characters. The most remark- able of such disruptions with which I am acquainted is that proposed by Rafinesque, in 1840, m "The Good Book and Amenities of Nature." With a religious dread of large genera, this author divided Carex into eighteen genera, and raised them, together with four genera made from Uncinia, into ordinal rank under the name Carexides. Long before this time he had divided the genus into four genera: Carex, Scuria, Triplma, and Triodus. In 1844, J. Heuffel, in Flora, adopted nine genera, of which eight were erected upon those species,

60 PROCEEDINGS OF THE AMERICAN ACADEMY

comparatively few ia number, wliicli lie without the division com- monly recognized as Carex proper. The genus Carex as received by Linnteus included Uucinia, which was separated by Persoon in 1807. In 1819, Beauvois, in Lestiboudois's " Essai sur la Famille de Cype- racees," proposed the genus Vignea, to include most of the distigma- tous homostachyous species, choosing the name in honor of Prof. G. F. de la Vigne, translator of Schkuhr's " Riedgriiscr " into French.

The sectional divisions of the genus have been built heretofore largely upon artificial groups. So far as I know, the monostachyous species have always been thrown together, until an attempt was made to distribute them in natural groups in Coulter's Manual of Rocky Mountain Botany. I have neglected such names as Unispicatce, MonostachyjB, Homostachyos, and Heterostachyas, as unsuitable for the designation of natural sections, however valuable they may be for artificial keys. For the same reason, I have not made use of the Dontostoma3 and Cyrtostomae of Fries, nor the Chlorostachyae and Melanostachyag of Tuckerman. In 1835 Elias Fries made a number of names, mostly plurals of the names of well-known and representative species, to designate some of the lesser groups of the genus. These were published in " Corpus Florarum Provincialium Sueciae." This idea was followed to a small extent by Kunth, in 1837, in the second volume of " Enumeratio Plantarum." In 1843, Prof. Edward Tuckerman published his curious and critical " Enume- ratio Methodica Caricum quarundam," a pamphlet of twenty-one pages, which was (he first professed attempt to make a natural ar- rangement of Carices with named divisions. A year later Drejer's excellent " SymbolEe Caricologicae " appeared, in which the general affinities of many species were discussed at length, and eleven sections proposed for the true Carices. The next important additions to the names of minor groups were made by John Carey, in the first edition of Gray's Manual, 184S.

SUBGENUS I. EUCAREX, Cosson, Fl. Paris, 744. Staminate flowers forming one or more terminal linear or club-shaped spikes (which, however, are often pistillate at base or apex). Pistillate flowers usually in distinct and simple mostly peduncled spikes. Cross- section of the perigynium circular, obtusely angled, or prominently trigonous in outline. Style mostly 3-parted and the achenium trigo- nous or triquetrous.

Section I. PHYSOCARPiE, Drejer, Symb. Car. 10 {Deflexo- carpoi, Bailey, Coulter Man. 373, in part). Perigynium mostly

OF ARTS AND SCIENCES. 61

straw-colored at maturity, papery in texture, more or less inflated, smooth, nerved, tapering into a beak as long or longer than the body : spikes few to many, distinct, compactly flowered : stigmas mostly three. — The representative species of the section are the larger members of the Vesicaria?, and Lupulinte. The extreme is represented on the one hand by tlie monostachyous C microglochin and G. pauci- jiora, and on the other by the comose and green-spiked C Pseudo- Cyperus. But even with these widely dissimilar extremes the section is a natural one. There are complete and almost insensible grada- tions from the one limit to the other. Most of the Lupulinas and the extreme species of the Pseudo-Cyperse do not have straw-colored perigynia until full or over maturity, while the perigynia of the Pauciflora3 and Pseudo-CyperEe are scarcely inflated or papery in texture. Occasionally the nerves are indistinct, rarely wanting. C Grayii alone has hispid perigynia, and that rarely. The species of this section, almost entirely North American and European, are mostly large and stout, and probably to be regarded as the most developed of the genus.

A. Paucijlorce, Tuckermaii, Enum. Meth. 7. (Leucoglochin, Fries, Summa, 73. Orthocerates, Koch, Fl. Germ. 748, is a sectional name made for V. microglo- chin.) Perig3nium greenish, linear-lanceolate or ahiiost needle-shaped, not inflated, strongly deflexed at maturity, several times longer than the incon- spicuous scale : spike androgynous, the pistillate flowers at the base, few. — Small species, rare or local.

1. Carex microglochin, Wahl. Kongl. Acad. Handl. xxiv. 140.

Uncinia microglochin, Spreng. Syst. Veg. iii. 830 ; C. B. Clarke,

Journ. Linn. Soc. xx. 401. Uncinia Europcea, J. Gay, Flora, 1827, 28. Remarkable for the elongated rhacheola which projects from the perigynium, completely filling the orifice. This plant stands midway between Carex and the singular genus Uncinia. — Colorado, Hall & Harbour 607 ; Greenland, Andersson. N. Europe, Alps, Himalayas.

2. Carex pauciflora, Lightfoot, Fl. Scot. 543, t. 6.

Cold swamps : Vermont, Central and "Western New York and Central Michigan, northward and northwestward to N. Minnesota, Sandberg, Rocky Mts. of British America, Drummond, and Sitka, Bongard, Mertens.

B. LupulincB, Tuckerman, Enum. Meth. 1.3. Perigynium green or greenish- tawny or sometimes yellow, more or less inflated (except in C. subulata), long, usually very turgid at the base, mostly erect or nearly so, very gradu- ally lengthened into a long slenderly toothed beak, exceeding the scale:

62 PEOCEEDINGS OF THE AMERICAN ACADEMY

spikes three or more, the staminate mostly one and commonly stalked, the pistillate often sessile, usually short and thick, often becoming dark colored in drying. — Mostly large and coarse species in meadows and bogs.

* Plant green, in appearance muck like those of the last group, very slender ; peri-

gynium needle-shaped with refiexed teeth, not inflated.

3. Carex subulata, Michx. FI. Bor.-Am. ii. 173.

G. ColUnsii, Nutt. Gen. N. Am. PI. ii. 205. G. Michauxii, Dewey, Sill. Joum. x. 273. Deep cedar swamps in sphagnum : Canada, Michaux ; Rhode Island, Olney, to New Jersey, Bewey, Nuttall, Carey, Parker ; Schuylkill Co., Penn., Porter; Fayetteville, N. Carolina, Curtis ; Aiken, S. Carolina, Ganhy^ and Georgia, Neisler. Rare.

* * Whole plant yellowish ; staminate spike sessile or nearly so ; pistillate spikes loosely

flew flowered ; perigytiium long-lanceolate, more or less spreading at maturity, somewhat turgid.

4. Carex Michauxiana, Boeckeler, Linnsea, xli. 336.

G. rostrata, Michx. Fl, Bor.-Am. ii. 173, not With. G. xant/iophysa, vars. nana and minor, Dewey, Sill. Joum. xiv. 353, ff. 57, 58. In cold bogs: Newfoundland, La Pylaie., to mountains of New England and Eastern New York ; Northern Michigan, Porter. Local.

5. Carex folliculata, Linn., Sp. Plant. 978.

G. xanthophysa, Wahl. Kongl. Acad. Handl. xxiv. 152. Scales, at least the lower ones, rough-awned and nearly as long as the. perigynium. Leaves broad and flat. — Cold swamps : Newfound- land, La Pylaie, to New England, New York, Pennsylvania, New Jersey, Parker, and Michigan, Wheeler and Smith's Cat.

Var. AUSTRALIS.

G. floUiculata, Ell. Sk. Bot. ii. 545 ? Chapm. Fl. 544.

G. flolliculata, ^, Boott, 111. 91.

Plant much smaller and more slender, with narrower bracts and

leaves ; pistillate spikes longer and looser ; perigynium much more

slender, scarcely inflated, more spreading ; scales, at least all above

the lowest, much shorter than the perigynium, muticous or slightly

awned. — Florida and northward, Chapman; New Orleans, Herb.;

St. Augustine, Florida, Canhy ; " dam^) pine land," Santee Canal,

South Carolina, Ravenel.

* * * Plant green ; staminate spike commonly stalked; pistillate spikes thick and compactly flowered ; perigynium very turgid at base.

6. Carex intumescens, Rudge, Linn. Trans, vii. 97, f. 3.

C. folliculata, Wahl. Kongl. Acad. Ilaudl. xxiv. 152, fide Boott.

OP ARTS AND SCIENCES. 63

Varies considerably in the size of the spikes. — Common in moist, shady pastures and in swamps : Norway House, S. W. of Hudson's Bay, about lat. 54°, Herb.; Newfoundland, La Pylaie ; throughout the States east of the Mississippi. Evidently more common north- ward.

7. Carex Gratii, Carey, Sill. Journ. iv. 22.

O. intumescens, var. globularis, Gray, Ann. N. Y. Lye. iii. 236. Perigynia sometimes hispid (see Coult. Bot. Gaz. x. 295). — Central New York to Central Michigan (common) and Illinois ; New Jersey, Brinlon, Closter, Austin ; Rome, Georgia, Chapman. Rare eastward.

8. Carex lurida, Wahl. Kongl. Acad. Handl. xxiv. 153; Fl. Lapp. 250.

O. lupuUna, Muhl. ; Willd. Sp. PI. iv. 266.

C. Canadensis^ Dewey, Sill. Journ. 2d ser. xli. 229. Very variable in the shape and size of the spikes. — Hudson's Bay, Boott ; common in wet places in the Northern States east of the Missis- sippi, rarer southward ; " deep river swamps, Florida and northward," Chapman; Santee Canal, South Carolina, Ravenel ; Apalachicola, Florida, Chapman ; Decatur, Alabama, J. D. Smith ; Limestone Gap, Indian Terr., Butler ; Houston, Texas, Lindheimer.

Var. DIVERGENS.

C. Bella-villa., Dewey, Sill. Journ. 2d ser. xli. 229. Plant more slender ; spikes scattered or remote, oblong or cylindri- cal, much more loosely flowered, more or less staminate at the apex ; perigynium more straw-colored, less turgid, slenderly beaked, diverging at right angles ; scales longer, conspicuously awned. Much like C.fol- licidata, from which it may be distinguished by habit, nari'ow leaves, longer (1 inch or more) spikes, the upper of which are sessile. Pos- sibly a hybrid with C . follicidata. — Belleville; Canada West, Macoun.

Var. POLYSTACHYA. '

C. lupidina, var, polystachya, Schwein. & Torr. Monogr. 337.

C glgantea, Kunth, Enum. PI. ii. 503.

C. lupuliformis, Sartwell,.Exsicc. 147.

C. Beyrichiana, Boeckeler, Linnoea, xli. 239. Penn Yan and Jefferson Co., New York, Sartwell, Crawe, to Connecticut, Wright, and New Jersey, Carey, etc., and Delaware, Canhy ; Georgia, according to Boeckeler, I.e.; Red River, Louisiana, Hale ; " Fort Smith to Rio Grande," Bigelow. Runs into the species in Michigan and other central States.

64 PROCEEDINGS OF THE AMERICAN ACADEMY

9. Carex gigantea, Rudge, Linn. Trans, vii. 99, f. 2. Distinguished from C. lurida, var. polystachya, which it closely

resembles, by its more slender and spreading pistillate spikes, its less turgid, more abruptly beaked and spreading pei-igynia, and shorter and smooth scales. Staminate spikes one to five. — Swamps: Ken- tucky, Short; Delaware, Canby, Commons, to Florida, Chapman, and Texas, Hall, 758. " Pine barren ponds, P'lorida to South Carolina and westward," Chapman. Evidently local.

10. Carex Halei, Carey, Chapm. Fl. 543.

C. turgescens, Dewey, Sill. Journ. 2d ser. iii. 356.

G. Halei, var. minor, Boott, 111. 94. Banks of the Apalachicola River, Florida, Chapman, to Louisiana, Hale, Joor ; Arkansas, Carey, according to Boott, and E. Texas, Wright.

11. Carex Elliottii, Schwein. & Torrey, Monogr. 357.

C. castanea, Ell. Sk. Bot. ii. 546.

C. Baldwinia, Dewey, Sill. Journ. xxvi. 107. " Boggy margins of pine-barren streams, Florida to North Carolina," Chapman. I have seen specimens from Florida collected by Chapman, and from North Carolina collected by M. A. Curtis. A rare and pretty species.

C. Veaicnrice, Tuckerman, Enum. Meth. 13. {TentaculatcB, Tuckerman, 1. c.) Perigynium smooth and shiny, much inflated, at maturity straw-colored or occasionally purple, beaked and conspicuously short-toothed, usually promi- nently few-nerved, mucli shorter than in the Lupulinffi : staminate spikes commonly two or more : pistillate spikes as a rule long and densely cylin- drical. — The types of the group are C. vesicaria, C. monile, and C. Tiicker- tnani.

* Plants tall but slender ; spikes Jew-Jiowered, globular or short-oblong, Jew, scattered or remote, straw-colored; leaves long and narrow,

12. Carex turgescens, Torrey, Monogr. 419.

" Pine-barren swamps, Florida to North Carolina," Chapman. I have examined specimens from Florida collected by Chapman, from Society Hill, South Carolina, collected by M. A. Curtis, and from New Orleans in Herb. Torrey. Rare.

13. Carex oligosperma, Michx. Fl. Bor.-Am. ii. 174.

C. Oakesiana, Dewey, Sill. Journ. xiv. 351. Borders of swamps and lakes : Bear Lake, Arctic America, English River, and Norway House (lat. about 54°), Richardson, southward to N. Minnesota, Central Lower Michigan, Central New York, and Pennsylvania ; Labrador, Allen.

OP ARTS AND SCIENCES. 65

14. Carex Raeana, Boott, Rich. Arc. Exped. ii. 344.

Differs from C. oligosperma chiefly as follows : Pistillate spikes cylindrical, long {\ in. to 1^ in.), the lower loosely flowered at the base and peduncled ; perigynium long-beaked, conspicuously toothed, the teeth scabrous ; pistillate scales narrow and acuminate. — Methye Portage, British America, long, about 110°, lat. about 57°, Richard- son. Some of the perplexing and immature specimens from N. Maine, collected by G. E. Smith (Gray, Manual, 602), and from New Bruns- wick, collected by /. Fowler, are evidently to be referred here. The specimens vary in the shape of the orifice of the perigynium. Abun- dant and mature specimens from these regions are needed for the full determination of this species, and likewise of C. rotundata and O. saxatilis, var. miliaris. G. Raeana evidently occurs in Montana also. The species was named for Capt. John Rae, of Richardson's Arctic expedition.

15. Carex physocarpa, Presl, Reliq. Ha;nk. i. 205.

The only typical specimens which exist in this country, so far as I know, I have from Professor Macoun. The species is allied to G. monile on one side and G. saxatilis on the other. Its distinguishing marks are the long peduncles (1 to 4 in.) of the oblong spikes, the two or more elevated staminate spikes, and a very slightly toothed or obliquely cut nearly nerveless perigynium, which is about equalled by the somewhat obtuse scale. G. saxatilis, var. Grahami, is appar- ently its nearest ally, but that variety has shorter and thicker pistillate spikes which are short-peduncled, one or two short-stalked staminate spikes and a sharply cut, strongly nerved perigynium which is twice longer than the scale. The merits of G. physocarpa and G. saxatilis, var. Grahami, cannot be determined until more material accumulates. My present impression is that expressed by Dr. Boott (Linn. Trans. xix. 220), thai the plants of our Rocky Mountains, which have been referred to the var. Grahami, belong rather to G. physocarpa. The plants vary widely from the type of G. physocarpa, but fully as widely from var. Grahami. It may be that the two species are not distinct. Var. Grahami is founded upon a plant of Scotland. — Nootka Sound, Vancouver's I., Haenke ; Rocky Mountains of British America, Drum- mond, named by Dr. Boott ; Bow River at Morley, Rocky Mountains of British America, Macoun.

* * Plants mostly loiv, with short and more or less purple spikes. Perigynia erect or ascending.

16. Carex saxatilis, Linn. Fl. Lapp. 259.

G. pulla, Gooden. Linn. Trans, iii. 78.

VOL. XXII. (n. S. XIV.) 5

66 PROCEEDINGS OF THE AMERICAN ACADEMY

C. vesicaria, van alpigena^ Fries, Mant. ii. 142. Greenland, Vahl, Warming & Holm, Fries. Kamtschatka, Wright. N. Euro^ie.

Var. (?) Grahami, Hooker & Arnott, Brit. Fl. 8th ed. 510.

C. Grahami, Boott, Linn. Trans, xix. 215.

C vesicaria, var. dichroa, Anderss. Cyp. Scand. 18.

G. saxatilis, var. major, Olney, Bot. King's Rep. 370. See C phjsocarpa. — High mountains from Colorado and Utah northward.

Var. (?) MiLiARis, Bailey, Coult. Bot. Gaz. ix. 120. C. miliaris, Michx. Fl. Bor.-Am. ii. 174. (7. pulla, var. (?) miliaris, Gray, Manual, 5th ed. 602.

Very slender, 6 to 16 inches high; leaves and bracts very narrow, almost filiform : spikes 1-3, small (two lines to six lines long, two lines to three lines broad), sessile or the lowest very short-stalked, brown-and-green, the upper usually ovoid or globular, and sometimes very much reduced in size : perigyniura ovoid, small, nerveless or nearly so, little inflated, the beak minutely toothed, about the length of and broader than the acute purple-margined scale. — Moosehead Lake, Maine, Smith, and northward to Lower Canada and New Brunswick, Fowler. More material is needed. Singular specimens from Ungava Bay, N. Labrador, coll. by Turner, 1884, appear to belong here. This plant is said by Olney to occur in N. Minnesota.

-t- -1- Perirjijnia squarrose.

17. Carex compacta, R. Brown, Ross's Voy. Appx. cxliii.

C. memhranacea, Hook. App. Parry's 2d Voy. 406.

C hymenocarpa, Drejer, Revis. Crit. Car. 58.

C ampullacea, var. borealis, Lange, Rink's Groenl. ii. 118. Six to sixteen inches high, stout : culm smooth or very nearly so, longer than the flat leaves : lowest spike subtended by an abruptly spreading leafy bract : pistillate spikes usually two, mostly closely sessile (the lowest sometimes very short-stalked), densely and evenly cylindrical, in typical specimens about an inch long, often shorter : staminate spikes one or two, short and obtuse or nearly so : perigynium broadly ovate, very short-beaked, the orifice nearly entire, a little longer than the white-pointed scale, bladder-like and shining, very fra- gile. Evidently too near the next. — Arctic America : " Cumberland House [about lat. 54° and southwest of Hudson's Bay] to Arctic roast," Boott; Baffin's Bay, R. Brown ; Southampton I., Capt. Parry; North Somerset, Duckett Cove, Ross ; Kotzebue's Sound, Arnott ;

OF ARTS AND SCIENCES. 67

Greenland, Vahl, fide Drejer ; Nottingham I., Hudson's Straits, Ma- coun ; Bear Lake, Drummond ; Kamtschatka, Wright. Figured ia Bot. Gaz. for August, 1885.

18. Carbx rotundata, Wahl. Fl. Lapp. 235.

Distinguished from the last by its firmer perigynium and involute- filiform leaves. Andersson regards it as an extreme form of G. ros- trata, With. (C. ampuUacea, Gooden.). — Arctic America : Slave Lake and Fort Enterprise, Boott ; Greenland, Vahl, Warming & Holm. Specimens from Ungava Bay, N. Labrador, Tamer, 1884, are evi- dently this species. One of G. E. Smith's specimens from Moosehead Lake, Maine, may belong here also. N. Europe.

* * * Plants mostly large and stout.

t- Perigynium not conspicuously turgid, sguarrose at maturity and the spikes

comose in appearance.

19. Carex rostrata, Withering, Arrang. Brit. PI. ed. ii. (1790).

G. obtusangula, Ehrh. Calam. Exs. no. 50 (1791).

G. ampuUacea., Gooden. Linn. Trans, ii. 207 (1792). Mackenzie's River, Herb. ; Saskatchewan, Bourgeau, Macoun ; Ore- gon, Lyall, and others ; Colorado, Ganhy., Hall & Harbour 615, and others ; N. Michigan, Porter. Probably generally distributed through- out British America, the Rocky Mountain region, and westward. Europe.

Var. UTRICULATA.

G. utriculata, Boott, Hook. Fl. Bor.-Am. ii. 221.

G. utriculata, var. minor, Boott, 1. c. ; Sartwell, Exsic. no. 153.

G. ampuUacea, var. utriculata, Carey, Gray's Man. 1848, 566.

G. utriculata, var. globosa, Olney, Bot. King's Rep. 374. In bogs entirely across the continent north of Ohio. G. Bongardiana, C. A. Meyei', ex Trautv. & Meyer, Fl. Ochot. 101 (C. vesicaria, Prescott in Bongard's Obs. Sitcha, 169), is evidently an attenuated and long-beaked form of G. ampuUacea. Dr. Boott, how- ever, regarded it as a form of G. vesicaria. — Sitka, Mertens ; Oregon, Hall.

•^ -^ Perigynia conspicuously turgid, ascending at maturity. ^

20. Carex monile, Tuckerman, Enum, Meth. 20. G. Vaseyi, Dewey, Sill. Journ. xxix. 347.

Common in wet places in all the Northern States east of Nebraska ; Colorado, Vasey ; Ostrander's Meadow, California, Bolander 6211; " Cumberland House to Bear Lake," Boott.

68 PROCEEDINGS OP THE AMERICAN ACADEMY

Var. OBTUSISQUAMIS.

G. vesicaria, y, W. Boott, Bot. Calif, ii. 252. G. vesicaria, var. (?) obtusisquamis, Bailey, Carex Cat., and Bot. Gaz. ix. 121. Spikes short (an inch long or less), sessile, purple ; perigynium ab- ruptly contracted into a short nearly entire beak, longer than the broad purple and white-margined very obtuse scale. — Soda Springs, head of Tuolutnne River, California, Brewer, 1781.

21. Carex vesicaria, Linn. Sp. PI. 979.

Huntington Valley, Nevada, Watson 1270, a young specimen ; Cal- ifornia, Summit Camp, Dr. Kellogg, Yosemite Valley, Breioer 1654, Tomales Bay, Bolander 2303, Sprague River, east of Klamath Valley, Hildebrand ; Northern British Columbia, Rotlirock. Evidently rare. Europe.

Var. MAJOR, Boott, Hook. Fl. Bor.-Am. ii. 221.

Vars. lanceolata and globosa, Olney, Proc. Am. Acad. viii. 407,

408.

Mostly larger : perigynium long-lanceolate, greenish or rusty, little

inflated, thick in texture, many-nerved, much longer than the small

scale. — Oregon, Hall 608, 609, Henderson, Multnomah Co., Howell;

Columbia River, Douglas, Scouler.

22. Carex Tuckermani, Boott, Hook. Journ. Bot. v. 73. G. bullata, Dewey, Sill. Journ. ix. 71.

G. Tuckermani, var. cylindrica, Dewey, Sill. Journ. xlix. 48.

G. cylindrica, Carey, Gray's Man. 184*8, 566. Perigynia more inflated than in any other American species. — "W. New England and N. Vermont to Ohio, Illinois, and Wisconsin ; N. Jersey, Britton Cat.

23. Carex bullata, Schkuhr, Riedgr. Nachtr. 85, f. 166. G. cyUndrica, Schw. An. Tab.

G. Greenii, Boeckeler, Flora, 1858, 649. Culm very sharply angled, at least above, and slender. The peri- gynia have a peculiar greenish-brown appearance. The roughness of the beaks is not always apparent. — Wet places, Mass. to N. Jersey and Pennsylvania, Porter ; Society Hill, S. Carolina, Gurtiis.

C. BULLATA X UTRicuLATA (G. Olneyi, Boott, 111. 15, t. 42). — Providence, R. Island, Olney. Sterile or nearly so.

24. Carex retrorsa, Schweinitz, An. Tab. G. reversa, Sprengel, Syst. Veg. iii. 827.

OP ARTS AND SCIENCES. 69

Marshy places : River du Loup, Quebec, Pickering ; New England to Pennsylvania, Michigan, and Wisconsin ; Sault St. Marie, N. Michi- gan, Porter ; Saskatchewan, Bourgeau ; Northwest Coast, Douglas ; Oregon, Hall.

Var. Haktii, Gray, Man. 5th ed. 600.

C. Hartii, and var. Bradleyi, Dewey, Sill. Journ. 2d ser. xli. 226. C. Macou7ii, Dewey, Sill. Journ. 2d ser. xli. 228. Pistillate spikes loosely flowered, long, all scattered or remote, the lower ones on very long peduncles : perigynium spreading or somewhat reflexed. — Pompet, Vermont, Morgan ; New York, Dundee, Yates Co., Samuel Hart Wright, Ludlowville, Tompkins Co., Lord, Greece, near Rochester, Bradley ; Flint, Michigan, D. Clarke ; Hastings Rood and Belleville, Canada West, Macoun.

25. Carex tentaculata, Muhl. ; Willd. Sp. PI. iv. 266. a rostrata, Willd. Sp. PL iv. 282.

C. tentaculata, var. rostrata, Pursh, Fl. 14. G. Purshii, Olney, Exsicc. fasc. i. no. 30. Variable. — Common in wet grounds throughout the States east of the Mississippi ; Texas, Wright. South America.

Var. GRACILIS, Boott, 111. 94.

Smaller and slender, with about two small erect pistillate spikes which are densely cylindrical (an inch to an inch and a half long), the size of a slender lead pencil. — Mountains of Vermont and New Hampshire ; Tennessee, Lesquereux. One of the handsomest of our Carices.

C. TENTACULATA X LURiDA (var. ? altior of C. tentaculata, Boott, 1. c). — Penn Yan, N. York, Sartwell ; Amherst, Mass., Tuckerman; North Hero, Vermont, Morong.

D. PseudocijpercE, Tuckerman, Enum. Meth. 13. Perigynium less inflated, more conspicuously nerved or even costate, provided with more or less setaceous or aristate teeth ; scale usually aristate ; spikes mostly nodding, comose in appearance, in color greenish, greenish-yellow, or whitish-yellow.

26. Carex Schweinitzii, Dewey, Sill. Journ. ix. 68.

A fine local species. — " Canada to New Jersey," Boott ; " New England, New Jersey, W. New York, and northward," Gray's Manual ; Pittsfield, Mass., and Pownal, Vermont, Dewey ; New York, Cowles, Fairfield, Gray.

27. Carex hystricina, Muhl. ; Willd. Sp. PI. iv. 282.

C Cooleyi, Dewey, Sill. Journ. xlviii. 144.

70 PROCEEDINGS OF THE AMERICAN ACADEMY

C. Georgiana, Dewey, Sill. Journ. 2d ser. vi. 245.

C. Thurberi, Dewey, Bot. Mex. Bound. 232. Wet meadows and marshes throughout the States east of the Mis- sissippi from Canada to Georgia, but evidently inclining to be local ; New Mexico, Wright ; Santa Rita Mts., Arizona, Pringle. — The name was originally written hyslericina, a name of no application. That the author meant to refer to the comose or hystricinous character of the spikes is evident from Willdenow's German name of the plant, no doubt suggested by Muhlenberg, " Stachelschweinartige Segge," porcupine-like sedge.

28. Carex acutata, Boott, Linn. Trans, xx. 124.

C.feminea, Steud. Cyper. PI. 203. Much like the last, fi'om which it differs in its greater size (two to four feet high) : pistillate spikes long and thick (one and a half to three inches long), erect, the lowest one or two short-stalked: peri- gynium elliptic-ovate, thick in texture, dull, very prominently many- ribbed, abruptly contracted into a rather short, white nerveless beak, longer than the rough and aristate scale. — 8. Arizona, near Fort Huachuca, Lemmon 2903 (1882). Founded upon South American specimens.

29. Cakex Pseudo-Cyperus, Linn. Sp. PI. 978.

Swamps and lakes : New England to Pennsylvania and Michigan ; Portage de Ratz at Fort de la Riviere, Winnipeg, Bourgeau. Evi- dently not common. Europe, Asia, Africa, South America. Var. COMOSA, W. Boott, Bot. Calif, ii. 252. C.furcata, Ell. Sk. Bot. ii. 552. C. Pseudo-Cyperus, Schw. & Torr. Monogr. 355. C. comosa, Boott, Linn. Trans, xx. 117. Common east of the Mississippi, in low grounds, from Canada to Georgia ; Multnomah Co., Oregon, Howell ; near San Francisco, Cal., Bolander 2301.

E. Squarrosce, Carey, Gray's Man. 1848, 564. Perigynium obconic or obovoid, squarrose in exceedingly dense spikes.

* Spikes mostly three to six inches long.

30. Carex spissa.

Culm four to six feet high, very stout, from a woody root, smooth or nearly so : leaves very numerous, a half-inch wide, stiff, glaucous, conspicuously serrate on the margins, about as long as the culm : lower bract very long and leafy, the uppermost short or nearly obso- lete: spikes 6-12 or more, the lowest four to six inches long and long-

OF ARTS AND SCIENCES, 71

peduncled (peduncle one to fifteen inches long), the upper becoming sessile, often two or three from an axil, the intermediate ones often staminate at the top, the pistillate portions densely and evenly cylin- drical, all erect, the staminate four to six or more in number and one to four inches long ; perigynium small (a line and a half long), yellowish green, elliptic or obovate, firm in texture, few-nerved, often slightly compressed, squarrose, about the length of or shorter than the stout and toothed awn of the scale : stigmas three. — California, San Diego Co., Pringle^ San Juan Capistrano, J. C. Nevin ; Arizona, Pringle ; Lower California, Guadeloupe Canon, Orcutt. The stoutest Carex I know.

* * Spikes not exceeding an inch and a half in length.

31. Carex stenolepis, Torrey, Monogr. 420.

C Frankii, Kunth, Enum. PI. ii. 498.

C. Shortii, Steud. Nomen. Bot. 296. Swamps and meadows ; Central Pennsylvania, Porter, to the upper districts of Georgia, westward to Kentucky, Short, and southwestward to Louisiana and Texas, Wright, Reverchon.

32. Carex sqdarrosa, Linn. Sp. PI. 973.

C. typhina, Michx. Fl. Bor.-Am. ii. 169. C typhinoides, Schwein. An. Tab. Bogs, New England to Georgia, and west to Michigan and Illinois; N. W. Ai'kansas, Harvey ; Red River, Louisiana, Hale.

Section II. TRACHYCHL^NiE, Drejer, Symb. Car. 9. Pe- rigynium mostly thick and hard in texture, commonly scabrous or hirsute, straight-beaked ; pistillate spikes compactly flowered, mostly large, erect or nearly so ; staminate spikes one or more ; stigmas three. — Large and coarse species, mostly paludose. A heterogeneous section, originating in the Anomalce with very short and entire beaks and terminating in the much-developed beaks of the Paludosae. The Anomalae connect the section with the Granulares. C vestita, C. Ore- gonensis, and O. Houghtonii are not paludose, and perhaps others of the section are not. C Jiliformis and C acutiformis may be said to represent the section.

A. Shortiance. Terminal spike androgynous, staminate below ; perigynium small and smooth, nearly beakless, entire. — Including one anomalous species, which Dr. Boott prefers to associate with C. verrucosa, EU.

33. Carex Shortiana, Dewey, Sill. Journ. xxx. 60.

G. Shortii, Torr. Monogr. 407.

C formosa, Kunth, Enum. PI. ii, 431.

72 PROCEEDINGS OF THE AMERICAN ACADEMY

S. Pennsylvania to Virginia and westward to Oliio, S. Illinois, Nebraska according to Dewey, and Limestone Gap, Indian Territory' Butler. ^'

B. ^«omate, Carey, Gray's Man. 1848, 557. Terminal spike all staminate • pistillate spikes long and cylindrical, mostly densely flowered; perigynium broad and short, short-beaked, the orifice very slightly notched or entire mostly granulate. — Tall species with rough leaves.

34. Carex Joori.

Culm two feet high, very sharply angled, rough: leaves narrow, slightly carinate, rough on the margins, much surpassing the culm : bracts sheathless, setaceous, very rough on the margins, the lower one or two surpassing their spikes, the others much shorter : spikes about six, cylindrical, densely flowered, one or two inches long, erect or slightly spreading, all on slender rough peduncles an inch or two long, the terminal distant and staminate : perigynium short-obovate, nearly circular in cross-section, abruptly contracted into a short entire and sharp beak half as long as the body, strongly many-nerved, granu- late, somewhat inflated, dark-colored, squarrose, about as long as the serrate awn of the hyaline scale: achenium triquetrous, broldly ob- ovate or oval ; stigmas three. — Comite Swamp, near Baton Rouge, La., /. F. Joor, Aug. 5, 1885.

35. Carex scabrata, Schweinitz, An. Tab.

Transition to the Granulares. — New England to South Carolina, J. D. Smith, and Tennessee, Curtiss, and westward to Michio-an.

36. Carex amplifolia, Boott, Hook. Fl. Bor.-Am. ii. 228, t. 226. Culm stout (two to three and a half feet high), very sharply

angled, rough above : leaves many, very broad (5-8 lines), rough on the margins, usually exceeding the culm: bracts leaf-like, sheath- less: pistillate spikes four to six, erect, the upper ones sessile, the lower very short-stalked, all erect, one to four inches long, slenderly cylindrical: perigynium elliptic, conspicuously few-nerved, abruptly narrowed into a whitish beak, spreading, about as long as the awn-pointed whitish narrowly purple-margined scale. — California, Plumas Co., Mrs. BidweU, Mariposa Grove, Bolander 5011 ,• Oregon! Hall., Howell, etc. ; N. Idaho, fide Boott.

C. /T/r/rr, Tuckerman, Enum. Meth. 14. (Lrrs/owr/.rt', Fries, Summa, 70. Lanu- ginosm and Scariosm, Carey, Gray's Man. 1848, 500.) A heterogeneous group distinguished from the last by the longer and more deeply cut beak and by the hairy perigynium. — The perigynium of C. striata is often smooth. C. filifonnis and C. IJougktonn may be taken as types of the group.

OP ARTS AND SCIENCES. 73

* Beak entire or nearly so.

37. Carex Sartwelliana, Olney, Proc. Amer. Acad. vii. 396. Whole plaut softly pubescent : culm two to three feet high, rigid :

leaves about two lines broad, about the length of the culm : bracts leafy, sheathless : stamiuate spike one, sometimes bearing a few pistil- late flowers : pistillate spikes four or five, sessile, an inch or two long, densely flowered, ferrugineous : perigynium trigonous-obovoid, mi- nutely beaked, the orifice entire, densely tomentose, a little longer than the mucronate purple-margined and ciliate scale. — Named in memory of Dr. Henry P. Sartwell of Penn Yan, New York, an ardent botanist, an early student of this genus, and author of an Ex- siccattE. — Yosemite Valley, California, Brewer 1636, Bolander 6221.

38. Carex vestita, Willd. Sp. PI. iv. 263.

In sandy soils. New England and New York to Pennsylvania, Porter, and Georgia. " Sandy swamps in the upper districts," Chapman.

39. Cauex hirtissima, W. Boott, Bot. Calif, ii. 247.

Culm rather slender but strict, a foot and a half high, smooth : leaves shorter than the culm, pubescent, those on the culm producing conspic- uous pubescent sheaths : spikes three or four, the upper short-stalked, erect, about an inch long, scattered rather loosely flowered, usually ex- ceeded by the leafy bracts : perigynium triangular-elliptic, contracted at both ends, nerveless or nearly so, hirsute, ribbed on the angles, pro- duced abruptly into a slender beak, about the length of the white and very abruptly ai'istate scale. The beak becomes more or less bifid by splitting at full or over maturity. — Summit Camp, Sierra Nevada, California, Dr. Kellogg.

* * Beah distinctli/ bifid.

40. Carex Oregonensis, Olney, Proc. Amer. Acad. 1872, 407. C. Halllana, Bailey, Bot. Gaz. ix. 117.

Culm a foot high, smooth or nearly so, very leafy : bracts leaf- like, with thin white auricles, the lower exceeding the culm : pistillate spikes about three, an inch long, often staminate at the top, erect, ap- proximate, shortly peduncled : perigynium ovate, tapering at both ends, prominently many-nerved, thickly covered with short stiff hairs, grad- ually produced into a white and smooth toothed beak, longer than the ovate, acute, membranaceous, and often dull-margined scale : achenium large, triangular-obovoid or rarely lenticular-obovoid. — Name given by Mr. Olney in the list of Hall's Oregon plants without description, and the numbers of the collection became mixed. In the Olney

74 PROCEEDINGS OF THE AMERICAN ACADEMY

herbarium at Providence I found the plant with Mr. Olney's name

attached. Orecfon, Hall 605, Howell, sandy hillsides on subalpine

slopes of Mt. Hood, L. F. Henderson, 1884 ; Washington Terr., Suksdorf.

41. Carex striata, Michx. Fl. Bor.-Am. ii. 174.

C.polymorpha, Torn Monogr. 413. New Jersey to Florida in pine-barren swamps. " Quite common in pine-barren regions, and mostly confined to the yellow drift," Britton's N, Jersey Catalogue. The Southern specimens have more hairy and more tapering perigynia than the New Jersey form.

42. Carex Houghtonii, Torrey, Monogr. 413.

Dr. Douglas Houghton, a brilliant naturalist, first State Geologist of Michiiran. — New Brunswick, Fowler ; Maine, at Orono, Scrib- ner, Milford, and Cumberland, Blake ; Gilmanton, N. H., Blake ; New York, near Lake Placid, Essex Co., Peck; shore of Lake Ontario (in New York?), Whitney, 1849; Michigan, Clare Co., central part of lower peninsula, Wheeler, Keweenaw, Bobbins, and Isle Royal, Por- ter ; sixty miles north of Belleville, Ontario, Macoun : Minnesota, Lake Itasca, Houghton (the original station), Blue Earth Co., Leiberg; Council Bluffs, Geyer ; British America on Athabasca Plains, Macoun, Saskatchewan, Herb., Methye Portage, long, about 110°, lat. about 57°, Richardson. Rare.

43. Carex filiformis, Linn. Sp. PL 976.

In bogs from New England to Pennsylvania and Michigan, and northwestward to the Saskatchewan and northward. Europe. Var. LATiFOLiA, Boeekeler, Linnaja, xli. 309. C. lanuginosa, Michx. Fl. Bor.-Am. ii. 175. a pellita, Muhl. ; Willd. Sp. PI. iv. 302. C. cemathorhyncha, Olney, Bot. King's Rep. 373. G. filiformis, YSiV. cBmatorhyncha, W. Boott, Bot. Calif, ii. 250. In bogs across the continent from Ohio and Kentucky northward ; also in New Mexico, Wright.

44. Oarex hirta, Linn. Sp. PI. 975.

Resembles C. Houghtonii and G. trichocarpa, but differs in its very remote and smaller pistillate spikes and its loosely hairy perigynium, sheaths, and leaves. — Introduced at Ashland, Mass., Morong, where it is thoroughly established, about Boston, W. Boott, in ballast at Phila- delphia, Scribner, and at Ithaca, N. York, Dudley.

D. PaludoscE, Fries, Corp. 190. (Lacustres and Aristata;, Carey, Gray's Man. 1848, 561.) Staminate spikes two or more, long-stalked; pistillate spikes

OF ARTS AND SCIENCES. 75

two to several, usually all peduncled, long and heavy, loosely flowered, erect or nodding : perigynium large, thick in texture, strongly nerved, mostly smooth, usually produced into a long beak which terminates in conspicu- ous awl-like erect or spreading teeth. — Coarse species.

45. Carex trichocarpa, Muhl. ; Willd. Sp. PI. iv. 302.

G. striataj Carey, Gray's Man. 1848, 561. C. trichocarpa, var. tiirhinata, Dewey, Sill. Journ. xi. 159. Marshes throughout the States east of the Mississippi, south to Georgia.

Var. iMBERBis, Gray, Man. 5th ed. 597.

Perigynium smooth, teeth usually shorter, pistillate scales longer and sheaths scabrous. — New York, Sartwell, to Illinois. Bolander's no. 4689, referred here in Bot. Calif, ii. 251, is probably not this plant.

Var. Deweyi, Bailey, Bot. Gaz. x. 293.

C. Iceviconica, Dewey, Sill. Journ. xxiv. 47. Big Sioux and Yellowstone Rivers, Hayden, Bismarck, Dakota, Seymour, and northward.

Var. aristata, Bailey, Bote Gaz. x. 294.

C aristata, R. Br. Narr. Frankl. Exp. App. 764.

C. atherodes, Sprengel, Syst. Veg. iii. 828.

G. orthostachys, C. A. Meyer, Fl. Alt. iv. 231.

G. mirata, Dewey, Wood's Bot. 1848, 593.

G. aristata, var. longo-lanceata, Dewey, Sill. Journ. xviii. 102. Geuerally distributed from New England to Oregon, and far north- ward.

46. Carex Watsoni, Olney, Bot. King's Rep. 370.

Culm erect, a foot and a half high : spikes about seven, deep red- brown, the four uppermost staminate, the topmost much the longest : scales of the staminate spikes lanceolate and acute or aristate, hispid, red-brown on the margins, three-nerved and pale in the centre : pistil- late scales lanceolate and abruj^tly aristate, ciliate at the apex : peri- gynium deeply cleft, the teeth spreading and clothed with a few lax hairs. — Founded upon very young and unsatisfactory specimens. I imagine that it is G. riparia. — Washoe Mountains, on a creek bank at the mouth of King's Canon near Carson City, Nevada, alt. 4,500 ft., Watson 1246.

47. Carex serratodens, W. Boott, Bot. Calif, ii. 245.

Culm one to two feet high, smooth, leaves a line or two broad, mostly shorter than the stem : bracts leafy, the lowest exceeding the culm, sheathless or nearly so, with purple auricles at the base : spikes

76 PROCEEDINGS OP THE AMERICAN ACADEMY

3 or 4, the terminal club-shaped and staminate, an inch or so long, the others pistillate, an inch or less long, sessile : perigynium lance-ovate, rather obscurely nerved, smooth, the spreading teeth conspicuously his- pid, mostly longer than the ovate and awned brown scale. — Found by Wm. Boott in Herb. Olney, without date, collector, or locality, but ticketed "California."

48. Carex riparia, W. Curtis, Fl. Lend. iv. t. 60.

G. lacustris, Willd. Sp. PI. iv. 306. Throughout the States east of the Mississippi and in Louisiana ; Pend d'Oreille Kiver, N. Idaho, Lyall ; Arctic America, Herb, Europe.

49. Carex acutifonnis, Ehrhart, Calam. Exs. no. 30.

G. paludosa, Gooden. Liun. Trans, ii. 202. Thoroughly established at Dorchester, Mass. ; boggy meadow, New Bedford, Mass., Willey.

Section III. MICRORHYNCH^, Drejer, Symb. Car. 9. (In- cluding JErostachyce and Melananth(B, Drejer, 1. c.) Parallel with the last section : distinguished in general by the much smaller and nearly or entirely beakless and mostly entire-mouthed perigynium which is much thinner in texture : stigmas two or three. — Paludose and alpine species of various habit, mostly possessing colored spikes, often grow- ing in dense tufts or tussocks. This is the most critical section of the genus. It is largely represented in Europe and North America.

A. AtratoR, Kunth, Enum. PI. ii. 431. Terminal spike club-shaped and andro- gynous with the staminate flowers below ; pistillate spikes mostly short and dark-colored, either erect or drooping ; plants mostly alpine or boreal. — The members differ widely in habit, the spikes of C. Mertensii and C. atrata, var. ovata, being mostly light-colored or parti-colored and drooping, while those of the others are black or nearly so, and more or less erect, although some black-spiked forms of C. atrata have drooping spikes.

50. Carex alpina, Swartz, Lilj. Svensk. Fl. ed. ii. 26.

G. Vahlii, Schkuhr, Riedgr. 87. G. media, R. Br. App. Frankl. Narr. 763.

G. alpina, var. nigrescens, Oluey, Prelim. Cat. Wheeler's PL 53. From Isle Royale, Lake Superior, and the mountains of Colorado and Oregon, to Arctic America. Europe. Asia.

Var. HOLOSTOMA.

C. holostoma, Drej. Revis. Crit. Car. 29. Differs from G. alpina chiefly in its beakless perigynium. — Green- land, Fries, Warming & Holm, etc.

OF ARTS AND SCIENCES. 77

51. Carex Mertensii, Prescott, Bong. Veg. Sitcha in Mem. Acad. St. Petersb. vi. ser. ii. 168.

Culm, two feet high, rough on the very sharp angles, very leafy : leaves short, broad, and flat, loosely and conspicuously sheathing, rough : bracts leaf-like, sheathless or nearly so, the lower exceeding the culm : spikes five to eight, an inch or two long, very densely flowered, all more or less staminate at the base, the upper one con- spicuously so, all drooping on filiform scabrous peduncles : perigynium very flat and very broadly ovate, winged, very thin in texture, green or whitish, few-nerved, appressed, much longer and broader than the purple white-ribbed scale : stigmas mostly 3. — Named for Karl Hein- rich Mertens, 1796-1830, of St. Petersburg, who made a botanical voyage around the world. — Marais Pass, Montana, Canhy ; Oregon, Cusick, Hall 599, Henderson ; Washington Terr., Lyall, and northward to Sitka, Mertens.

52. Carex atrata, Linn. Sp. PL 976.

White Mts., N. H., and high mountains of Wyoming, Colorado, Utah, Arizona, and California, and northward. Europe. Asia. Var. nigra, Boott, 111. 114.

G. nigra. All. Fl. Ped. ii. 267. Westward with the last. Europe. Var. ovATA, Boott, 111. 114.

C. ovata, Rudge, Linn. Trans, vii. 96, t. 9. Mountains of Colorado and Utah southward ; Newfoundland, Rudge, the original locality, but not since found there, so far as I know. Var. ERECTA, W. Boott, Bot. Calif, ii. 239. High mountains of Nevada and California.

53. Carex Gmelini, Hooker, Bot. Beechey's Voy. 118, t. 27. Differs from C. Buxbaumii, to which it is very closely related, by

its shorter and thicker spikes, which are ped uncled and darker colored, the ferrugineous and smooth (not granulated) perigynium and the hispid awn of the scale. — Named for Johann Georg Graelin, of St. Petersburg, 1719-55, author of Flora Sibirica. — Oregon, according to Boott, to Alaska, Herb. Siberia.

54. Carex Buxbaumii, Wahl. Kongl. Acad. Handl. xxiv. 163. Bears the name of Johann Christian Buxbaum, an early European

botanist, 1693-1730. — Bogs throughout, from Georgia northward and across the continent. Europe.

B. RigidcB, Fries, Summa, 72. ( Ccespitosce, Fries, 1. c. Stylosce, Bailey, Coulter's Man. 387.) Mostly stiff and rigid species, largely alpine or boreal, with

78 PROCEEDINGS OP THE AMERICAN ACADEMY

short and erect very closely flowered spikes, an entirely staminate terminal spike, and bracts possessing purple or black auricles at their base. — Species ranging from six to eighteen inclies high, rarely higher. The typical form of C. vuhjaris may be taken as the type of the group.

» Stigmas three, or in C. heteroneura perhaps often two.

55. Carex bifida, Boott ; Olney in Proc. Am. Acad, vii. 394. Much like C. Buxbaumii. Culm two or three feet high, sharply

angled, smooth or nearly so : leaves two lines broad, pale, mostly shorter than the culm : bracts narrow, sheathless, the lowest leaf-like and exceeding the culm: pistillate spikes two to five, very short and thick (usually about a half-inch long), sessile and contiguous, some- what glaucous : perigynium triangular-ovate, pale green, punctate, conspicuously nerved, the orifice cleft and the teeth ciliate, longer and broader than the purple white-ribbed, obtuse or muticous scale. Ter- minal spike rarely bearing a few pistillate flowers above. — California: Salinas Valley, in rather dry soil, Brewer 574 ; Pacheco Pass, Santa Clara Co., Bolander 4837 ; Red Mountains, Humboldt Co., Bolander 6476; Yreka Co., E. L. Greene.

56. Carex heteroneura, W. Boott, Bot. Calif, ii. 239o

Culm slender but erect, nearly two feet high, sharply angled, sca- brous or smooth : leaves narrow (two lines or less broad), flat, shorter than the stem : lower bract leaf-like, nearly as long as the culm, sheathless : pistillate spikes small (about a half-inch long and three lines or less broad), sessile or the lower ones on very short peduncles, more or less contiguous, whitish or yellowish : perigynium oval, smooth, straw-colored, few-nerved or nerveless, thin in texture, pro- duced abruptly into a very short beak which is slightly emarginate, longer and broader than the obtuse or muticous purple white-ribbed scale. If the terminal spike were androgynous the species would •strongly resemble C. atrata, var. erecta, although the perigynia are narrower and not so flat as in that variety. Variable in the nerving of the perigynia, — California, in the Sierras, Bolander; from Lake Talioe to Bear River, Kellogg ; Co3ur d'Alene Lake, N. Idaho, Watson 437.

57. Carex Raynoldsii, Dewey, Sill. Journ. xxxii. 39.

C. Lyallii, Boott, 111. 150, t. 483. Named for Capt. W, F. Raynolds, of an early government explor- ing expedition. — Mountains from Wyoming and Montana to Cali- fornia and Washington Territory.

58. Carex Parryana, Dewey, Sill. Journ. xxviii. 239.

C. arctica, Dewey, 1. c.

a Hallii, Olney, Hayden's Rep. 1871, 496.

OP ARTS AND SCIENCES. 79

Bears the name of Capt. Parry, the Arctic explorer. — South Park, Colorado, Wolf, Hall & Harbour 617, and northward to Hudson's Bay. Rare. 59. Carex stylosa, C. A. Meyer, Act. Acad. St. Petersb. i. 222,

t. 12.

C. nigritella, Drejer, Revis. Crit. Car. 32. Culm slender but erect, a foot or a foot and a half high, scabrous : leaves narrow (scarcely exceeding a line or a line and a half), mostly sliorter than the culm : staminate spike one, an inch or less long, slender, very short-stalked: pistillate spikes two or three, a half-inch or less long, the lowest on a slender peduncle an inch or less in length, the others sessile or nearly so : lowest bract nearly or quite equalling the culm : perigynium turgid-ovate, fuscous, minutely punctate, nerve- less, the entire orifice closed with the stiff and persistent style from which the stigmas are caducous, longer than the very obtuse black white-ribbed scale, A pretty species. — Fox Harbor, Labrador, Allen ; Greenland, Vahl ; Sitka, Bongard, Mertens.

Var. viRENS.

Stouter : leaves broader : spikes all closely sessile and much thicker: perigynium green, much broader than in the species and the scales much narrower. Perhaps a good species. — Sauvie's Is- land, Oregon, and Mt. Adams, Washington Territory, at 5,000 feet, Howell.

60. Carex Tolmiei, Boott, Hook. Fl. Bor.-Am. ii. 224. C. vulgaris, Olney, Proc. Am. Acad. 1872, 407. Culm rigid, a foot to foot and a half high, triquetrous, smooth or nearly so : leaves ordinary, rough on the margins, mostly shorter than the culm : lower one or two bracts leafy and about equalling the culm, sheathless : spikes four to six or seven, the uppermost an inch or less long, staminate and mostly short-peduncled, the others mostly con- tifuous, oval or oblong (three fourths inch or less long), dark-colored, often staminate at the apex, the two or three lowest on slender pedun- cles, an inch or two long, the others sessile : perigynium compressed- trigonous, oval or oval-oblong, pale and more or less discolored with purple dots, lightly nerved, produced into a very short and entire cylindrical beak, either longer or shorter than the obtuse or muticoua purple white-ribbed scale. — Tolmie was an Indian trader and early botanical collector. — Columbia River, Tolmie; Oregon, Hall 597: Northwest Coast, according to Boott ; Behring Straits, Langsdorff. Kamtschatka, Wright.

80 PROCEEDINGS OP THE AMERICAN ACADEMY

61. Carex nigella, Boott, Hook. Fl. Bor.-Am. ii. 225.

Smaller than the last: staminate spikes usually more than one: perigynium elliptical, the orifice bidentate : scales lanceolate and mucronate. — Columbia River, Tolmie.

* * Stigmas two, or very rarely a specimen may bear a few flowers with three

stigmas.

G2. Carex c^spitosa, Linn, Fl. Suec. ed. ii. 333.

a Facifica, Drejer, Fl. Ex. Hafn. 292.

C. Drejeri, Lang, Fl. Ratisb. 548. Differs from G. vulgaris in its always single staminate spike, its short pistillate spikes (three fourths inch or less long), which are closely sessile and contiguous, and its round or ovate, gibbous, entirely nerveless, somewhat squarrose, pale perigynium. — Said by Boeckeler to occur in British America, and by Drejer to occur in Greenland. Specimens collected by Dr. Gairdner in Oregon and by Dr. Richardson on the arctic coast were referred here by Dr. Boott, but the specimens are too young for satisfactory identification. No. 593 of Hall's Ore- gon collection, distributed as G. ccespitosa, is G. acuta, var. pallida. Europe.

Var. FiLiFOLiA, Boott, 111. 182.

G. aperta, var. angustifolia, Boott, Hook. Fl. Bor.-Am. ii. 218.

Culms weak, often with basal sheaths somewhat reticulated : leaves weak and slender, more or less involute, about as long as the culm : spikes dark-colored, slender : perigynium more or less ovate, obscurely nerved, deciduous. — Fort Good Hope, Mackenzie's River, lat. 67°, Dr. Richardson; Cascade Mts., lat. about 49°, Dr. Lyall, a stouter form. The materials are far too scanty to allow of a satisfactory estimate of any of the American representatives of this species.

63. Carex vulgaris, Fries, Mant. iii. 155.

G. ccespitosa, Gooden. Linn. Trans, ii. 195, t. 21.

G. Goodenovii, Gay, Ann. Sci. Nat. 2d ser. xi. 191.

G. anticcensis, Kunze, Suppl. Riedgr. 50, t. 13.

G. turfosa, Fries, Summa, 228.

G. Umula ? Gray, Man. 5th ed. 582. Newfoundland, Thaxter, to New England and Wisconsin ; Twin Lakes, Colorado, Wolf ; Alaska, Herb.

Var. JUNCELLA, Fries, Summa, 230.

G. Kelloggii, W. Boott, Bot. Calif, ii. 240. Wahsatch Mts., Utah, Watson, Jones; Union Co., Oregon, Cusick, and in British America, Macoun.

OP ARTS AND SCIENCES. 81

Var. HYPERBOREA, Boott, 111. 167.

C. Bigelovii, Torr. ; Schwein. An. Tab.

C Washingtoniana, Dewey, Sill. Journ. x. 272.

C. saxatilis, Dewey, Wood's Bot. ed. i. 581.

C. saxatilis, var. Bigelovii, Torr. Monogr. 397.

C. hyperborea, Drej. Revis. Crit. Car. 41. ^

C. rigida, var. Bigelovii, Tuckerman, Enum. Meth. 19.

C. limula, Fries, Summa, 229.

a dnhitata, Dewey, Wood's Bot. 1861, 755. White Mountains to Greenland ; Mountains of Colorado and N. California to Alaska.

Var. ALPiNA, Boott, 111. 167.

C. rigida, Gooden. Linn. Trans, ii. 193, t. 22.

C. saxatilis of Scand. authors, not Linn. Westward with the last.

Var. BRACTEOSA.

Culm slender, a foot or more high, rough, erect, bearing a long and leafy auricled bract about equalled by the flat pale leaves : pistillate spikes two or three, small and globular, black or nearly so, sessile, the one or two upper ones adjoining the small staminate spike (which is a half-inch or less long), the lowest one sometimes remote and borne in the axil of the long bract : perigyniura much as in var. alpina, but smaller. — Ebbett's Pass, California, alt, 8,000 feet. Brewer 2015.

64. Carex decidua, Boott, Linn. Trans, xx. 119.

C. Andersoni, Boott, Hook. Fl. Antarct. ii. 364. See Bot. Gaz. x. 204 (plate). — Differs from C. indgaris as follows : spikes heavier : scales and perigynia deciduous : perigynium conspicuously stipitate and strongly nerved. — California, Thurber, according to Boott ; Oregon, Herb. Founded upon specimens from Terra del Fuego and the Falkland Islands.

65. Carex nudata, W. Boott, Bot. Calif, ii. 241. Distinguished from C. vulgaris by its fibrillose sheaths and decidu- ous perigynium, and from both that species and C. decidua by its long and thin finely punctate, lightly nerved perigynium, which is empty in the upper half. Not <