November 3, 1898] 



NA TURE 



^,, fi.,...H :». and on p. ijo/i'x) has been printed instead of 

 \x). On p. 73 the phrase "en outre" is an imperfect 

 equivalent for "wir setzen noch," and is likely to make 

 the reader suppose that the symbol * (.i', |) has been 

 already defined. The handsome appearance of the 

 volume, and the excellence of the printing, fully maintain 

 M. Gauthier-X'illars' high reputation. 



The first edition of the "Lehrbuch" was reviewed in 

 this journal at considerable length, so that it is un- 

 necessary to give here any detailed account of the con- 

 tents of this volume. Prof. Weber has introduced 

 various improvements in detail, and added, amongst 

 other tilings, an account of Lagrange's interpolation 

 formula, and of Hurwitz's very interesting researches on 

 Sturm's theorem. The extreme value and originality of 

 the treatise become more evident the more carefully it is 

 studied ; the appearance of this translation, as well as 

 that of the second edition at home, indicates that its 

 great merits are being duly appreciated. G. B. M. 



A Miiniicil of the Grasses of iVe'W Soutli Wales. By J. 



H. Maiden. (Sydney : Gullick, 1898.) 

 Broi'i.ht out under the authority of the Minister for 

 Mines and Agriculture for New South Wales, the 

 Government botanist publishes a \ery useful account 

 of the grasses of the Colony. The number of species 

 indigenous to the Colony is stated at ig6, comprised in 

 56 genera. These numbers may be compared with the 

 95 species belonging to 48 genera reckoned by Hooker 

 as natives of these islands. Under each species, in 

 addition to the technical diagnosis, the vernacular names 

 are given, with a reference to the published figures ; the 

 habitat and range of the species, and an account of its 

 value for fodder and for other purposes. Then follows a 

 key to the genera, and under each genus a key to the 

 species. In the case of a number of the more useful or 

 more common species, full-page illustrations are given. 

 The volume is a very useful and valuable one ; but, 

 considering its purpose, it strikes us that its practical 

 value would be increased by either a general description 

 of the flower of grasses, or a glossary of technical terms ; 

 unless indeed, Australian farmers are much better 

 acquainted than our own with botanical terminology. 

 Manuel de V E.xploratetir. By E. Blim and Rollet de 



risle. Pp. vii + 260. (Paris : Gauthier-Villars, 1899.) 

 The object of this volume is to provide travellers in little- 

 known regions with information which will enable them 

 to record particulars of service to geography concerning 

 the land traversed. In the first chapter the methods are 

 described for determining and representing approximately 

 the route followed and the details of the surface, using a 

 prismatic compass and an aneroid. The astronomical 

 observations required to define positions along the route 

 are then explained, and it is shown how the combination 

 of these observations with the rough survey enables an 

 approximately accurate representation of the journey to 

 be laid down. The determination of heights and dis- 

 tances by levelling and triangulation are described in the 

 third chapter, for the instruction of explorers who wish 

 to make a detailed study of particular districts. The two 

 lemaining chapters deal with systems of projection for 

 the conversion of the observations to maps, and the 

 choice and transport of the instruments referred to in the 

 text. The sextant is not included among the instruments, 

 the astronomical observations being made with the 

 theodolite instead. No instructions are given as to what 

 to observe in natural history, geology, anthropology, or 

 other sciences ; hence, the volume is not to be compared 

 in value with the " Hints to Travellers " published by the 

 Royal (.Geographical Society. Nevertheless, as a clear 

 and very elementary manual on surveying and practical 

 astronomy for travellers who explore without having re- 

 ceived a preliminary scientific training, the book may 

 prove of service. 



NO. 1514, VOL. 59] 



LETTERS TO THE EDITOR 



[ Tlie E'iitor does itol hold himself responsible for opinions ex- 

 pressed hy his correspondetils. Neither can he undertake 

 to re/iirtj, or to cor> espond with the writers ofy refected 

 iiiaiiiisc! i^'s intended for this or any other part of NATtJRE. 

 A'o notice is talcen of anonynioits cornnmnications.'X 



Heredity and Fertility. 



As Prof. Karl Pearson in his paper "On the Law of 

 Ancestral Heredity," published in the Proceedings of the Royal 

 Society, allutJes to an investigation which he has apparently 

 commenced into the inheritance of fertility in man, it may be of 

 interest now to publish one which I completed some years ago, 

 but which for various reasons has not appeared in print. 



The problem to be solved was : — Do marriages of heiresses 

 prove more or less fertile than ordinary marriages ? The first 

 desideratum was of course to get a practically homogeneous class 

 from which to obtain the necessary data. The families enrolled 

 in " Burke's Peerage and Baronetage" sufficed. Taking the 

 volume for 1S92 — just published when this investigation was 

 commenced — there were in it of the present generaUon 265 

 ordinary or non-heiress marriages which had produced an 

 average of 3'S children per marriage, forty-eight of which, or 

 iS'5 per cent., were absolutely sterile. 



Now an heiress, so called in '• Burke," may be considered of 

 the same class as an only child. Hence, taking care that both 

 parents lived long enough after the birth of their first child to 

 have had others, I extracted all the marriages of heiresses and 

 female only children that occurred in " Burke " for the present 

 generation, with the addiiion of a considerable number from the 

 past generation contained in "The Lineage." There resulted : — 

 243 marriages, averaging 374 children per marriage and 1 1 '9 

 per cent, marriages absolutely sterile. These figures show that 

 there is practically no difference in the fertility of ordinary 

 marriages— 3'S per marriage— and the 374 fertility of heiresses 

 ar.d only children : the small difference of six-hundredths per 

 cent., on so small a number as 250 marriages, being well within 

 the ordinary limits of error. It implies that with greater numbers 

 the results would probably be similar. Not .so, however, when 

 we come to consider the number of marriages which are 

 absolutely sterile, for these are so much fewer with the only 

 children marriages — 6 '6 per cent.— that we must say when they 

 are fertile they are more fertile than the others. 



But the problem may be attacked from the other side. If 

 female members of the smallest families possible tend to bear 

 (ewer children than the female members of families of ordinary 

 size, the female members of large families would tend to bear 

 large families. Is this so ? With a much greater amount of 

 labour than would be imagined, I extracted from the same work 

 some marriages of ladies who were members of families of five 

 or more, i.e. had in each case four or more brothers or sisters, 

 with the result that 250 marriages averaged 4"25 per marriage 

 with 1 3 '6 marriages quite .sterile. 



Summarising, we get then Udies of small families average 

 4 06 children per marriage, 11 'I per cent, marriages being quite 

 sterile. 



Ladies of ordinar)- families average 3'8 per marriage, lS'5 

 quite sterile. 



Ladies of large families average 4'25 per marriage, I3'6 

 quite sterile. 



This shows that the more contrasted the ladies are as 

 regards number of brothers and sisters, the more nearly allied 

 are the numbers of their children and the relative amount of 

 absolute .sterility. A conclusion tending to .show that the cause 

 under investigation i^ not a vera causa. 



Does the female influence show itself in a change in the pro- 

 portion of male to female children, assuming of course a chance 

 distribution of the husbands as being members of small, ordinary, 

 and large families? 



Sons Daughters. Total S. to loo D. Sterile. 

 243 heiress and only-child 



marriages 492 414 906 107 sons II 



265 ordinary marriages ... 514 52S 1042 97 sons iS'S 

 250 large f.imily marriages 553 510 1063 loS sons 13"6 



The same result as before — the most contrasted classes give 

 the most similar results. 



If heredity has anything to do with this matter, which so far 

 seems improbable, it may arise from the other side— the father's. 

 Hence I extracted marriages of only sons, of sons of ordinary 



