April 19, 1900] 



NATURE 



589 



LETTERS TO THE EDITOR. 



t The Editor does not hold himself responsible for opinions ex- 

 pressed by his correspondents. Neither can he undertake 

 to return, or to correspond with the writers of, rejected 

 manuscripts intended for this or any other part of NATURE. 

 No notice is taken of anonymous communications. '\ 

 The Eclipse-Wind. 

 Although meteorological observations during total eclipses 

 of the sun are of secondary importance, the changes of tempera- 

 ture and humidity being well known, yet it is still uncertain 

 whether those changes do occur in atmospheric pressure, and, 

 con<;equently, of wind, which, theoretically, ought to result from 

 the sudden chilling of the air by the passage of the moon's 

 shadow, and the consequent increased barometric pressure out 

 of which the wind should blow in all directions. 



The letter of Mr. J. W. Evans, in Nature of December 28, 

 1899, describing his observations during the Indian eclipse of 

 1898 is an interesting contribution to the subjectof the " eclipse- 

 wind," and brings to mind the special investigations undertaken 

 by Prof. Winslow Upton and myself during the total solar 

 eclipses of August 19, 1887, 'n Russia, and of January i, 1889, 

 in California, the results of which are described, respectively, in 

 Ainer. Meteor, yournal, vol. iv., and in Annals Astron. Ob- 

 servatory of Harvard College, vol. xxix. No. i. Aneroid 

 barometers, including a recording instrument, all with open 

 scales, were used, and, in addition to the anemometer, a record- 

 ing wind-vane was taken to California. There the sky was 

 clear, but in the Russian eclipse it was heavily clouded. As 

 regards the atmospheric pressure, it must be said that, while in 

 both eclipses minute rises occurred during the total phase, yet 

 they cannot be attributed with certainty to its influence, since 

 similar fluctuations occurred on other days. As regards the 

 wind, the eclipse appeared to produce an appreciable effect, for 

 the wind backed (contra-clockwise) before totality, and veered 

 (clockwise) to its original direction afterwards, its velocity 

 diminishing as the eclipse progressed. This is what would be 

 expected to happen at a station situated near the central path of 

 a shadow moving north-east. In both eclipses it was nearly 

 calm during totality. 



With the hope of settling these questions, the writer joined 

 the Harvard Observatory party that observed the total solar 

 eclipse of April 16, 1893, in Chile, where not only was clear 

 weather assured, but the regular diurnal period of the barometer 

 afforded an excellent opportunity to study any non-periodic dis- 

 turbance due to the eclipse. Besides the previous instruments, 

 Richard's " statoscope," or differential barograph, was em- 

 ployed, with which, when the temperature is kept constant or 

 allowed to change at a uniform rate, variations of pressure 

 approximating 0"025 millimetre (i/iooo inch) of mercury are 

 recorded. In order to secure a free exposure in all directions 

 for the observations of wind, the station was located on the 

 summit of a mountain. Notwithstanding perfect conditions — 

 the day being clear, and a counterpart of those preceding and 

 following — no unusual changes in pressure during any of the 

 phases of the eclipse could be detected, and if any variation 

 occurred, it was insufficient to disturb the regular diurnal period, 

 and must have been of the order of a thousandth of an inch. 

 The record of wind-direction again showed a backing of the 

 wind prior to totality, and a veering round afterwards ; but as 

 these oscillations were not infrequent at other times, they cannot 

 c ascribed certainly to the eclipse. The wind reached its mini- 

 num velocity soon after the first contact of the moon's limb, and 

 -leadily increased until after the fourth contact. 



While these observations seem to prove that any change in the 

 atmospheric pressure during a total solar eclipse is so small as to 

 escape measurement, yet there does appear to be evidence of 

 changes in the wind. Mr. Evans ob.served a contrary rotation of 

 the wind to that described above ; and the reports of the changes 

 in wind during many eclipses, which were collected by the late 

 Mr. Ranyard {Memoirs Roy. Astron. Soc. vol. xli. chap, xxxv.), 

 are very contradictory. 



Therefore, it would be interesting if, on May 28, observers 

 along the path of totality in the United States and elsewhere 

 would make frequent observations of the direction and strength 

 of the wind. Still more valuable data could be obtained from 

 a few self-recording wind-vanes and anemometers exposed high 

 enough above the ground to be free from local influences. 



A. Laurence Rotch. 

 Blue Hill Meteorological Observatory, U.S.A., April 3. 



NO. 1590, VOL. 61] 



Lord Kelvin's Origip of Granite. 



It ii a sound maxim that if you want a thing done you must 

 do it yourself. So, as no expert has replied to my query as to 

 the soundness of Lord Kelvin's theory of granite, propounded in 

 my letter of February 23, I have consulted a big Dana's 

 " Mineralogy" with the following results. 



It may be premised that Lord Kelvin assumes for his liquid 

 lava a specific gravity of 2*50, but as according to Dana the 

 basaltic lava of Kilauea is in one case as high as 3 '20, the 

 primeval liquid lava may have averaged 2 "60. 



Although 2 "50 will work out well enough, 2 "60 is much more 

 striking as an illustration of the effect of the assumed convection 

 currents upon volcanic minerals. 



The following is a list of volcanic minerals in the order of 

 their density, with a rough indication of their composition so far 

 as soda, potash, lime, magnesia and iron are concerned : — 



Labradoriie ... Lime ... 2 '67-2 76 



Amphiboles ... .Magnesia, lime, iron ... 2 -90-3 -40 

 Pyroxenes ... Magnesia, lime, iron ... 3"23-3'50 



Free silica is represented by tridymite 2 "28-2 "33, instead of 

 by the heavier quartz. 



If we take 2*60 to be the specific gravity of the primeval 

 liquid lava, the division between crystals that would float and 

 those that would sink comes between sanadine and labradorite. 

 Some little allowance should be made for expansion on heating. 



According to these specific gravities, it would appear that the 

 snow shower produced by the convection currents would not 

 have the effect of silting up the lava ocean with granitic crystals 

 set in a mother liquor of basalt, but would have the effect of 

 differentiating the lava into light and heavy strata, until the 

 convection currents themselves would be checked and the sur- 

 face stratum, composed largely of potash and soda silicates, left 

 free to freeze. And, we may note, that the upper stratum is 

 composed of the raw materials of granite, while the lower 

 stratum is composed of the raw materials of basalt. 



At this point geology and petrology commence work, and 

 what subsequently befals the primeval crust, after the advent of 

 water and sediment, may be read between the lines of the great 

 works of MM. Daubree, Fouque and Levy. 



The question of a floating crust affects no doubt the problem 

 of the age of the earth, but that is beside my point, which is 

 strictly confined lo the origin of granite. 



Torquay, April 3. Arthur Roope Hunt. 



Is New Zealand a Zoological Region? 



In your issue of January il, Mr. H. Farquhar wrote drawing 

 attention again to the incongruity of associating New Zealand 

 with Australia in a zoo-geographic sense. He correctly insists 

 that the New Zealand fauna is not most closely allied to that of 

 North-east Australia (Queensland). It is significant that those 

 writers who advocate the alliance of New Zealand to Queens- 

 land have not seen either country, while those who deny such 

 relationship have studied or travelled in both or either areas. 

 No observer who had a fir.st-hand knowledge of the two coun- 

 tries could agree with Dr. Sclater that " it is probable that the 

 whole fauna of New Zealand has been originally derived from " 

 Australia. 



In the following number (p. 273), Dr. A. R. Wallace, writing in 

 support of his own and Dr. Sclater's views, does not demonstrate 

 or reaffirm their accuracy, but merely lays stress upon the 

 inconvenience of an opposite view. 



That an error is convenient is no good reason for its mainten- 

 ance. Regardless of the direction in which they point, our first 

 care must be the accuracy of facts and deductions. 



But, as Dr. Wallace implies, there may be fairly laid upon 

 destructive critics the burden of restoring by constructive work 

 the effects of their ravages. "If," says Dr. Wallace, "antipodean 

 naturalists restrict the ' Australian Region ' to Australia and 

 Tasmania, what shall be done with the remainder of his own 

 Australian Region ?" I have proposed {Journ. Malacology, iv. 

 1895, p. 55) that New Zealand, New Caledonia and neighbouring 

 groups (inclusive certainly of the Solomons, perhaps of New 



