July 2 1, 1887] 



NA TURE 



267 



developments of the century. Strangely enough, there is 

 in the whole work no mention of that most widely- spread 

 of all electric inventions, the domestic electric bell, nor 

 of its almost forgotten inventor, John Mirand. Prof. 

 Mendenhall has added to the interest of his sketch by 

 supplying a number of illustrative cuts of objects of 

 historic interest, such as Faraday's first magneto-electric 

 machine, and his first transformer or induction-coil. We 

 should have welcomed some account of the great theorists, 

 Coulomb, Laplace, and Weber, who, with Sir William 

 Thomson and Clerk Maxwell, have, by their calculations 

 and mathematical developments, played so leading a part 

 in the progress of the century ; but the author would 

 probably have found it impracticable with the plan of his 

 sketch to deal with the labours of these intellectual giants. 

 In his less ambitious aim of popularizing the experimental 

 development of the subject he has succeeded admirably. 



OUR BOOK SHELF. 



The Fungus Hunter's Guide and Field Memorandum 

 Book., with Analytical Keys to the Orders and Genera, 

 illustrated, and Notes of Important Species. By W. 

 Delisle Hay, F.R.G.S. (London: Swan Sonnenschein, 

 Lowrey, and Co., 1887.) 



A FIELD guide and mentor is a welcome companion 

 for the practical botanist, provided it is so compiled as to 

 meet all the requirements of field work, otherwise it is 

 merely "a delusion and a snare." This little volume, 

 unfortunately, belongs to the " otherwise," for it is 

 insufficient, antiquated, and misguiding ; insufficient, 

 because it includes only a few species under each genus or 

 sub-genus, and these have been selected without manifest 

 reason ; antiquated, because, although dated 1887, it is 

 based upon the state of this branch of science in 1871, 

 and might have been published at that date, for all internal 

 evidence to the contrary ; and misguiding, because the 

 errors of 1871 are not corrected, the illustrative figures 

 are entirely without names of the species intended to be 

 represented, and more important or essential species are 

 excluded than many of those included in the lists. 



Under each genus or sub-genus in the volume a list is 

 given of "common or notable species," — each with its 

 scientific name (but without the authority for the specific 

 name, which any botanist would regard as essential) ; an 

 imaginary popular ■ name, which is useless because 

 imaginary and not real ; a short description, rarely suffi- 

 cient ; and letters indicating esculent or poisonous 

 qualities. As only one or two species are given under a 

 genus or sub-genus which has a dozen or more other 

 British representatives, it should have been stated dis- 

 tinctly that there are so many more species which are not 

 named, any of which the collector might meet with in his 

 rambles. Unfortunately the selection of the species 

 favoured with a place has been made with very little 

 judgment. Some are included which are so rare that they 

 have only been found once or twice in this country, whilst 

 others are excluded which are almost sure to be met with 

 in any moderately successful ramble. The fact is patent 

 that the " Hand-book " issued sixteen years ago is accepted 

 as the authorized record for to-day, whereas it is abso- 

 lutely out of date, and all the great advances made during 

 the intervening period are studiously ignored. The 

 volume is interleaved with ruled paper for notes and 

 memoranda, and we venture to affirm that this is the 

 only useful and unexceptionable portion of the work. The 

 purchaser must judge whether it would not have been 

 more economical to secure a blank memorandum book, 

 since the numerous figures are valueless without names, 



and the analytical keys ought to have been more accurate 

 and better constructed. M. C. C. 



My Hundred Swiss Flowers : with a Short Account oj 

 Swiss Ferns. By Mary A. Pratten. (London : W. H 

 Allen and Co., 1887.) 



This is a very unpretending book, and should be of con- 

 siderable service to beginners in botany who may wish to 

 carry on botanical studies among the Alps during the 

 month of July or early in August. The writer has selected 

 those Swiss flowers which seem to her " most remarkable, 

 most characteristic of the country, or most commonly 

 seen," and she is, of course, right in thinking that a great 

 many of them will be new to such as make a first visit to 

 the Alps. Her descriptions are clear and sufficiently full, 

 and the illustrations are very good. 



LETTERS TO THE EDITOR. 



[The Editor does not hold himself responsible for opinions 

 expressed by his correspondents. Neither can he under- 

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

 rejected manuscripts. IVo notice is taken of anonymous 

 communications. 



[ The Editor urgently requests correspondents to keep their 

 letters as short as possible. The pressure on his space 

 is so great that it is impossible othennise to insure the 

 appearance rjen of communications containing interesting 

 and novel facts,'] 



The Carnatic Rainfall. 



Mr. H. Blanford's authority is so deservedly high, that I have 

 had some hesitation in writing to controvert the conclusions he 

 has adopted in the paper published in N.A.TUKE of July 7 

 (p. 227), entitled "The Eleven- Year Periodical Fluctuation of the 

 Carnatic Rainfall " ; and to state my reasons for thinking that 

 there is no real validity in the arguments he uses in favour of 

 "the very high probability that the apparent undecennial 

 fluctuation is no chance phenomenon." 



Mr. Blanford brings forward a series of figures which show 

 the mean annual variation of the rainfall during twenty-two years, 

 at a number of stations in that part of Southern India locally 

 known as the Carnatic, from the mean annual rainfall for the 

 Carnatic generally. From these figures he has inferred the 

 appearance of two complete cycles of eleven years, with a 

 dominant periodical fluctuation. 



To test the character of this apparent peiiodicity he obtains 

 from these figures the two first terms of an harmonic expression 

 that shall represent the observed facts for an assumed eleven-year 

 period ; and he finds the mean difference between the observed 

 values and those calculated from the adopted harmonic expression 

 to be ±3 '5 inches, from which the mean probable error of any 

 of the calculated periodical values is found to be ± 070 inch. 



Now it is apparent that such a series of calculated values has 

 no physical signification whatever. The greater or less degree 

 of difiference between the observed and calculated quantities only 

 indicates how far the sums of the terms of the harmonic series 

 employed coincide with the series of observed quantities which 

 the calculated series was desi.ined to represent. It is also 

 obvious that by a sufficiently extended series of terms the cal- 

 culated quantities might be brought to agree, within any desired 

 degree of approximation, with those observed. No conclusion 

 whatever, therefore, can be based on the amount of the differ- 

 ences above alluded to, so far as any question of periodicity is 

 concerned, and the so-called "probable error" is merely an 

 arithmetical result of the particular form of calculation adopted. 



Mr. Blanford goes on to remark that the mean difference 

 between the observed series of values of the annual variation of 

 rainfall and the mean of the whole of them, is ±5 '2 inches, 

 with a probable error of the general average of ±0*94 inch. 



And here again I am unable to see that any weight can be 

 attached to these figures in connexion with the main point at 

 issue. The mean variation of the series of observed values, from 

 the mean of all of them, will of course be greater than the 

 mean variation of those observed values from a series deliber- 

 ately calculated so as to correspond with them, such as that 

 obtained by aid of the harmonic series. The introiluction of the 

 expression "probable error" of the general average is alsti 



