> 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 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 *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'-6 .....^