December 1, 1896.] 



KNOWLEDGE. 



265 



ILLUSTRATED MACAZINI 



[€E. LITERATI 



LONDON : DECEMBER 1, 1896. 



CONTENTS. 



PAGE 



Forecasting Tamines in India. By Dotglas Aechibald 265 

 Greek Vases— IV. C. — Red - Figured Vases. Bv H. B. 



AVai.tees, M.A., F.S A. (lUusirafed) {Plate) ..'. ... 267 

 Some Curious Facts in Plant Distribution. — VI. Br 



W. BoTTiNG Hemslkt, F.R.S .'. 270 



Our Fur Producers. -VI. Bv R. Lydekkee, B.A , F.E S. 271 

 Letters: — W. H. S. Mosck; E. Brown; E. E. Maekwice ; 



D. FLAyn'.Y; .r. M.Wadmoee; C. Robixsox; C.A.Schott 275 



Notices of Books 277 



A Possible Cause of Change on the Moon's Surface. 



Bv CiiAT!i,E3 Davison, Sc.D., F.Ct.S. 278 



The Lunar Metropolis. By E.W.:\rArN-DEE,F.R.A.S.(i'/a^e) 279 



Science Notes 280 



Waves. — XII. Earthquakes, Nerve Waves, and Tele- 

 pathy. By Vaughan C0ENI3H, M.Sc. (Illustrated)'... 281 



The Beech. By Geoege PAXTO>f. {Illustrated) 282 



Helium and Parhelium. Bv E. W. Mauxdee, F.R.A.S. 284 



Face of Sky for December. By H. Sadleb, F.R.A.S. ... 287 



Chess Column. By C. D. Locock, B.A.Oxon 287 



EDITORIAL NOTE. 



In completing the Volume for 1896, the Editors desire 

 to record their satisfaction at the favourable reception 

 accorded to the ilagazine in its extended area of work 

 during the year. It is intended to pursue the same course 

 in the new year, and while keeping abreast of the highest 

 scientific achievements and aspirations of the time, every 

 effort will be made to render the columns of Knowledge 

 of abiding interest to its readers. " Simply worded — 

 exactly described " will be the aim in the future as it has 

 been in the past. 



With the January Number will be commenced a series 

 of illustrated papers on the Progress of Science durinLr the 

 Queen's unique reign, a period of time unexampled in the 

 world's history for the scientific advancement of the 

 nations. Arrangements have been made to enlist the aid 

 of leading men in the various branches of science, and 

 no pains will be spared to render the series complete in 

 every particular. 



The same Number will contain the first of a series of 

 illustrated articles on Practical Entomology assisted by the 

 ^licroscope, to be contributed by Mr. Frederick Enock, 

 F.E.S., F.L.S., the well-known entomologist. 



During the year, Mr. G. F. Hill, M.A., will continue his 

 interesting contributions on Numismatics, dealing first 

 with English Medals. These papers will be illustrated 

 with full-page photographic plates. Mr. W. Botting 

 Hemsley, F.K.S., will also contribute a series of articles 

 on the Vegetable Products of our Australasian Colonies. 



In the Natural Sciences, although no long series of 

 connected articles, other than those mentioned above, is 

 contemplated, the Editors have promises of many valuable 

 papers from both old and new contributors. Mr. E. Walter 

 Maunder, F.R.A.S., who will continue to edit the astro- 

 nomical columns, has received several photographs from Dr. 

 Isaac Eoberts, I'.H.S., while many other friends have pro- 

 mised their aid. These, together with several drawings of 

 planets by well-known artists in this work, will be issued 

 monthly in the form of collotype plates as heretofore. 



FORECASTING FAMINES IN INDIA. 



By Douglas Archibald. 



THE scarcity which is now developing into famine 

 in India, caused by an unusually early withdrawal 

 of the monsoon, draws attention to the possibility 

 of forecasting such events. 



The crucial test and ultimate aim of every true 

 science is accurate prediction. 



In physical problems, when all the antecedent causes 

 and operative laws are known, it is usually easy to 

 predict the result. Where, however, as in meteorology, 

 both the causes and the laws are only partially known, and 

 where the intrusion of unforeseen factors and combinations 

 may occur both in time and space, prediction must at 

 present be based more or less on empirical relations and 

 analogy. If carried beyond the time limits of direct 

 observation, it must be of a broad and general character, 

 in direct proportion to the time covered by the forecast. 



So far, in England, and in most European countries, 

 weather prediction has never advanced beyond the twenty- 

 four hour limit, and is based almost entirely on an 

 examination of contemporaneous data and empirical 

 relations founded on past experience, together with a 

 modicum of rational law. In America, owing to the wide area 

 traversed by ephemeral weather changes, the time limit is 

 extended to thirty-six hours ; and in Australia, disturbances 

 reaching Western Australia are forecasted to arrive in 

 Victoria about three or four days after, but this is the 

 extreme limit. 



It is beginning at the wrong end to predict the particular; 

 but so long as the method is confined to what is taking 

 place within the area of observation, though the deductions 

 are made on empirical bases, they can hardly fail to reach 

 a certain percentage of accuracy, sufficient for practical 

 purposes. 



When we extend the time limit to several months instead 

 of hours, and, ignoring the minor fluctuations, take note 

 merely of the general average prevalent character of the 

 weather during the whole period, a marked departure has 

 to be made in the system employed. 



^Monthly averages take the place of hourly means ; con- 

 trasts and analogies between the conditions for six monthly 

 periods preceding and embracing the monsoon periods 

 have to be studied. The area of observation also has to 

 be extended beyond the hmits^of the district dealt with, 

 in order to determine the subsequent efi'ects of ascertained 

 general conditions prevailing in surrounding areas. 



On such principles the Indian six monthly forecast has 

 been founded, and its success has been sufficient to induce 

 the Indian Government not merely to grant the fands 

 necessary to establish fresh observations in the Persian 

 area, but also to arrange for the transmission of cablegrams 

 from Mauritius. By these means, the conditions which 

 prevail in the South Indian Ocean may be reported to the 

 office at Simla in time to be utilized for the forecast. 



Although within certain limits, the summer monsoon — 

 which bursts, after being ushered in by heavy thunder- 

 storms, about June 6th in Bombay, and arrives at its 

 northernmost limits some two or three weeks later — is a 

 tolerably regular phenomenon, it is not nearly so regular 

 both in time and quality as is commonly supposed. Its 

 date of arrival, for example, occasionally varies as much 

 as thirty days, while the amount of its attendant rainfall 

 has varied from a deficiency of six inches in 1808 to a 

 surplus of nine inches in 1893. Concentrated in one spot 

 this latter excess would equal two hundred and eleven 

 cubic milos of water. To give an idea of what such an 



