Sept. 27, 1877] 



NA TURE 



455 



has spent his labour on a work of such little real value. 

 He seems, indeed, to have had some misgivings as to its 

 shortcomings, and has largely borrowed from Huxley and 

 Martin's "Elementary Biology." Thus the student is 

 left as best he may to harmonize Thome's account of the 

 antheridium of Characea:, on p. 293, with Huxley's 

 independent description on p. 394, which is transferred 

 almost bodily. 



It can hardly be doubted that Mr. Bennett, with his 

 experience as a teacher, could have supplied us from his 

 own pen with a text-book which would have been much 

 more useful. 



OUR BOOK SHELF 



Natural Ccomeliy ; an Introduction to the Logical Study 

 of Mathematics, for the Use of Schools and Technical 

 Classes, with Explanatory Models. By A. Mault. 

 (London : Macmillan, 1877.) 



This is a good elementary text-book, founded on the 

 work by M. E. Lagout ("' Takimetry "), which we have 

 already noticed (NATURE, vol. xvi. p. 226). The ground 

 covered by the work before us is not quite so extensive 

 in one direction as that covered by Dr. Gwynne's trans- 

 lation ; but it has an introduction to pure geometry which 

 is likely to be of service to junior pupils. We are dis- 

 posed to think that some such practical training as that 

 indicated here, with the aid of the accompanying models, 

 and a shoit course of "practical" geometry would be a 

 capital thing for our junior pupils. Boys who are 

 exceedingly dull and stupid over their " Euclid" often, as 

 we have repeatedly seen, lake much interest in these con- 

 crete exhibitions of geometrical truths. The book has 

 been very carefully got out ; there are a few loose expres- 

 sions which might be improved. On p. 32 is the statement, 

 '• in equal circles equal arcs are those which have equal 

 chords," a distinction should be made between major and 

 minor arcs. Another trifling matter (but some boys 

 would at once notice it) is that some equilateral figures 

 are drawn on p. 33, which are not equilateral by scale. 

 There are two parts — geometry by sight, which treats of 

 the measurement of flat surfaces and of solids, and scien- 

 tific geomttry, or reasoning helped by sight. The latter 

 is concerned with the measurement of accessible and 

 inaccessible things and with the incommensurable (as the 

 circle, sphere, cylmder, and cone). We can recommend 

 the book for school nse. 



LETTERS TO THE EDITOR 



[Tht Editor does nut hold hityiself responsible for opinions expressed 

 by his corresponde7its. Neither can he undcr'tilct to return, 

 or to correspond with the writers of, rejected manuscripts. 

 No notice is taken of anonymous communuations. 



The EdUor urgently requests correspondents to kiep their letters as 

 short as posiible. The pressure on his space is so great that it 

 is impossible otherwise to ensure the app arance even of com- 

 munications containing interesting and novel facts.'\ 



The Cycle of Sun-spots and Rainfall 



May I venture to ask insertion fur the following remarks on 

 the cycle of sun-spots and rainfall? I frequently receive in- 

 qmrtts regarding llie meteorological aspects of the Indian faniine 

 and the prospects of Madra- during the coming monsoon. But 

 I btg to st.iie that the object 01 my invesiigations is not at 

 present to predict the luure, lut simply to ascertain the pa-t, 

 lacts. When we are quite sure of the data it will be time 

 enough to ap]>ly them. In order to secure a stable basis it lias 

 been necessary to work up a vast accumulation of meteorologicil 

 returns which have never been j^revioujly collated, and to make 

 further references to India, Germany, and America. Some time 



must still ebp?e before 'he results can be presented as a whole. 

 M ear.wl ile I should be obliged if you would give insertion to 

 the fatts with regard to two points which seem at present to 

 have a special interest to the public. 



In the firs' place, I think I may now safely say that the coin- 

 cidence of the cycle of r.dnfall with the sun-spot period is not 

 confined to Madras, but is common to various points around the 

 great ocean tract on which Madras lies. The three such points 

 of observation for which materials exist during the most con- 

 siderable period of years are Madras, Bombay, and the Cape of 

 Good Hope. I have made up tlie following table from the 

 monthly, and, when available, from the hou'Iy, rfturns of these 

 stations ; and as there are a few errors (probilily clerical) in a 

 return recently given for Bombay by the meteorological reporter 

 at Calcutta, I ought perhaps to add that these returns have been 

 tested by a trained computator in the Scottish Meteorological 

 office. The sun-spot column was worked out from an old list, 

 the only one available to me in India, and will hereafter be 

 revised from the more complete returns issued this year by Dr. 

 Rudolph Wolf The diflJerences do not, however, affect the 

 general results. My cycle of eleven years .starts back from 

 1S76 ; and the minimum group of my cycle, namely, the 

 eleventh, first, and second years, include all the years of mini- 

 iiium sun-spots from 1877 to iSio. It will be seen that the 

 coincidence in the cycle at Madras and the Cape of Good Hope 

 is very stront;ly marked, while that at Bombay is less so, and 

 somewhat lags behind the other two. An explanaiion exists for 

 this, but it would trespass too much on your space to enter into 

 that side of the question. 



Table \.—Eli- 



Years' Cycle of Rainfall and Sun-spots shown 

 in Periods of Two Years. 



The cyclic coincidence maybe tested in another way. If it 

 really exists there should be a well-marked minimum group at 

 the extremiiies of the cycle (in the eleventh, first, .and second 

 jearh), and a well-marked maximum group in the middle of the 

 cycle (the fifth and followirg years). The years on both sides of 

 the central maximum group, i.e., between it and the minimum 

 group at the two extremities, should yield intenr.ediate results, 

 and, when taken together, should form an equally well-marked 

 intermediate group. I therefore divided the cycle (so far as the 

 number 11 permitted) into three equal groups. The "minimum 

 group" is formed by the three series of years at the extremities 

 cf the cycle, which include all the years of minimum sun-spots 

 in this century from 1810 to 1S77. The "maximum group" 

 embraces the four central years from the true maximum jear of 

 tf e rainfall and sun-spot cycle (the filth) to the second maximum 

 in the sub-cycle of sun-spots in the eighth year. The "inter- 

 mediate group" consists ol the two series of years on boih sides 

 01 the central maximum gioup, namely, the third ard fourth 

 years on ihe one side, and the ninth and ttnth years on the other 

 side. 



