NA TURE 



[OcTOliKR I, 1896 



profits from their mines, in which profits Europeans have 

 as yet no great interests," &c. 



After this it is not surprising, as the translator informs 

 us in the preface, that the pubhcation of tlic book 

 "caused quite a stir in Cicrman circles." 



OUR BOOK SHELF. 



Crystallography for Beginners, witli an Appendix on the 

 use of t lie Blowpipe <ind the Determination of Common 

 Minerals {after the method of Dr. Alhin Weisbach). 

 By C. J. Woodward, B.Sc. Pp. 164. (London: 

 Simpkin, Marshall, Hamilton, Kent and Co., Ltd., 

 1896.) 

 In a preliminary chapter of this book the student is 

 taught how to prepare for himself, with due regard to 

 economy of purse, a set of models to be used in con- 

 nection with the various lessons. In the course of the 

 following 72 pages the constancy of the angles of 

 crystals, symmetry, notation, drawing of crystal forms 

 -and spherical projection, are in turn explained. The 

 physical properties of crystals are then briefly touched 

 upon, and in a last lesson mero-symmetry is discussed. 

 The appendi.x (55 pages) deals with a subject entirely 

 different from Crystallography, namely Determinative 

 Mineralogy, and is made up almost wholly of tables 

 drawn up after the manner of those of Dr. Weisbach. 

 The book contains numerous woodcuts in the text, and is 

 furnished with four plates, two of them consisting of 

 diagrams to be pasted on cardboard and used in the 

 construction of the aforementioned models. To each 

 lesson is appended a set of useful cjuestions relating to 

 the subject which has been discussed. Some of the 

 statements are wanting in accuracy : for instance, on 

 page 55 the student is told that "the symbols of all 

 planes in a zone ha\e two of their indices always in a 

 constant ratio," which is untrue ; and at times the 

 language is wanting in neatness and precision : still, if 

 the student is in the hands of a careful teacher, he will 

 be able to get much help from the book, and is not likely 

 to be led astray. 



By the Deep Sea; a Popular Introduction to the Wild 

 Life of the British Shores. By Edward Step. Pp. 322. 

 (London : Jarrold, 1896.) 

 The author of this little volume is already favourably 

 known by his popular books on wild flowers, &c., and the 

 present work will add to his reputation as a writer for 

 the non-scientific reader. The author's endeavour has 

 been to introduce to the seaside visitor a large number 

 •of the interesting creatures to be found on the rocks, the 

 sands and the shingle, and he claims to have written the 

 whole of the work in close contact with the objects he 

 describes — not only of cabinet specimens, but of the 

 living creatures under natural conditions. In his own 

 words : " There is not a line in the whole volume that has 

 not been written within a few yards of, and in full view 

 of the rocks." The twenty chapters into which the book 

 is divided are devoted to the sea and its shores, low forms of 

 life, sponges, zoophytes, jelly-fishes, sea-anemones, sea- 

 stars and sea-urchins, sea-worms, crabs and lobsters, 

 shrimps and prawns, some minor crustaceans, barnacles 

 and acorn-shells, "shell-fish," sea-snails and sea-slugs, 

 cuttles, sea-squirts, shore fishes, birds of the seashore, sea- 

 weeds, flowers of the shore and cliffs. The style of writing 

 is easy and attractive, and the text is further elucidated 

 by the insertion of a number of well-chosen, if somewhat 

 rough, illustrations from the works of P. H. Gosse, and 

 others which appear to have been specially drawn for the 

 work. Many a seaside holiday will be more fully and 

 permanently enjoyed by the study of this tastefully got- 

 up little book, the usefulness of which is increased by a 

 general and a classified index. 



NO. 1405, VOL. 54] 



LETTERS TO THE EDITOR. 



[7 lie Editor does no! hold himself responsible for opinions ex- 

 pressed by his correspondents. Neillier can he undertake 

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

 tnanuscrifts intended for this or any other part of Naiure. 

 No notice is taken of anonymous commuitications.'\ 



The Utility of Specific Characters. 



I HOi'El) that I might have held my peace on this subject. 

 Prof. Lankesler, however, complains, and not fur the first time, 

 that I have misrepresented, or at any rate niisuiiderstood him. 



I do not doubt his acquaintance with Prof. Weldon's work, 

 though he has allowed a long time to elapse befure criticising it. 

 I am glad that he regards it as "interesting and valuable." 

 But this is what he said about it in Natukk Un July 16 last : — 



" Such methods of attempting to penelralc the obscurity 

 which veils the interactions of the immensely complex bundle 

 of phenomena which we call a crab and its environment, appear 

 to me not merely inadequate, but in so far as they involve per- 

 version of the meaning of accepted terms and a deliberate 

 rejection of the method of inquiry by hypothesis and verifica- 

 tion, injurious to the progress of knowledge." 



It is quite true that Prof. LanUester has not said in so many 

 words that " Prof Weldon's investigation of (he crab's carapace 

 'does not satisfy the canons of scientific inquiry.'" But it 

 appears to me that this is a very mild way of putting what he 

 did say. 



I expressed the opinion that Prof WeldonV investigation did 

 rest on an hypothesis, and that this was subjected to verification. 

 Whether the hypothesis was reasonable and the verification 

 adequate is a matter on which Prof Karl Pearson and others 

 are entitled to form their own judgment. 



Kew, September 28. W. T. Thiselton-Dyer. 



I KEEl. grateful to Prof Karl Pearson for his luciil and 

 rational contribution to this discussion, in which it has sometimes 

 seemed to me that the main question was in danger of being 

 obscured by more or less irrelevant arguments. 



I pointed out in a letter to Nature, soon after the publication 

 of Prof Weldon's report Ijst year, that he had not, and had not 

 claimed to have, proved that there was a ditfurential or selective 

 death-rate in shore crabs, with respect to variations of their 

 frontal breadth. He showed that the curve of variation in larger 

 (and therefore presumably older) crabs was tlifferent from that 

 in smaller crabs. The departures from the mean were less. He 

 concluded, that if this difference were not due to growth-changes 

 it must be due to the death of crabs with extreme variations. 

 But on the other hand it had to be proved that the difference 

 was not due to growth-changes. Changes in the proportions of 

 parts are so common during growth in so many animals, that it 

 seemed to me much more likely that the difference discovered 

 by Prof Weldon was due to such changes than to a differential 

 death-rate. I understand that he has since been investigating 

 what he calls the law of growth in these crabs, but so far as I 

 know he has not published any further results. 



I am glad to find that Prof Karl Pearson's o|)inion concerning 

 the conclusions to be drawn from the evidence published by 

 Prof Weldon, entirely agrees with mine. It would be very 

 interesting to learn now whether Prof Weldon is able to settle 

 the question of the changes occurring in the growth of shore 

 crabs, and either to confirm or withdraw his suggested conclusion 

 that the difference he described was due to selective death-rate. 

 It would take a good deal of evidence to convince me that shore 

 crabs in which the frontal breadth differed sHghtly from the 

 mean, died in greater numbers than those in which it was nearer 

 the mean. But if the evidence is iorthcomini^, I am ready to 

 accept it. It seems to me that Mr. Thiselton-Dyer is inclined 

 to accept the conclusion before the evidence is f irthcoming. He 

 seems to have overlooked the other possible explanation of the 

 result, namely changes in the same crabs during growth. 



I also maintained in my letter last year, as I'rofs. Lankester 

 and Karl Pearson maintain now, that if a differential death-rate 

 were demonstrated, it would still be necessary to discover how 

 that death-rate was caused, what was the relation between the 

 character in question and the conditions of life which caused 

 individuals with certain variations of the character to die off. 



I do not profess to be a specialist in logic, but it seems to me 

 that the fallacy into which Prof Weldon lias fallen is that of 

 confounding the categories. He maintains that if a certain 



