January 24, 1901] 



NA TURE 



299 



some human purpose ? In the widest sense, therefore, all know- 

 ledge which is desired must be useful in some way and for some 

 purposes, and, even on the most narrowly "utilitarian" inter- 

 pretation, the useful is nothing else or more than what satisfies 

 desire — except that an attempt is made to restrict it to the satis- 

 faction of a very limited set of desires. An absolutely useless 

 method or study would be one that could not be worked in any 

 conceivable manner or for any conceivable purpose, i.e. it would 

 be not merely useless, \>\xt false. The proper defence, therefore, 

 of the so-called "useless" researches seems to me to consist in 

 showing in the first place that the context of a science requires 

 them, and in the second in pointing out that it has always, so 

 far, proved possible to find a directly practical application for 

 what is organically connected with a system of knowledge. 

 Corpus Christi College, Oxford. F. C. S. Schiller. 



The Field-mice and Wrens of St. Kilda and Shetland. 



In his notice of Messrs. A. H. Evans and T. E. Buckley's 

 " Fauna of the Shetlands " (Nature, May 24, 1900, pp. 75 and 

 76), your reviewer regrets that the authors are silent in regard to 

 the special characters of the Shetland field-mouse, in view of my 

 own recent recognition of a peculiar representation of this type 

 in St. Kilda. It may be interesting to point out that in a recent 

 paper on geographical and individual variation in Mus sylvaticus 

 and its allies {P.Z.S., 1900, p. 387), I found myself unable to 

 separate the Shetland field-mouse (specimens of which I had 

 fortunately been able to examine), at least at present, from that 

 of Great Britain generally. I would not, however, therefore 

 necessarily bind myself to follow your reviewer in his suggestion 

 that the comparative distinctness of the local forms of wren and 

 field-mouse may guide us in forming a decision as to the 

 relative periods during which St. Kilda and the Shetlands 

 have been separated from the mainland. So many factors seem 

 to be brought into play in the evolution of a local race or sub- 

 species that it is, I fear, unsafe to rely too much on such points, 

 and I have a strong suspicion that the influence of the environ- 

 ment has been too little taken info account by recent writers. 

 At all events the field-mouse of Iceland would, it might be 

 thought, show remarkable deviations from the mice of Western 

 Europe, yet the little that we know of it only shows us how 

 closely allied it is to Mus sylvaticus proper. 



As regards the wrens of Iceland and of the various Scotch 

 islands, attention may be directed to an interesting series of 

 measurements of wings given by Mr. R. M. Barrington in a 

 footnote to p. 641 of his book on " The Migration of Birds as 

 Observed at the Irish Lighthouses and Lightships." These 

 seem to show a gradual diminution in size from the large Tro- 

 glodytes borealis of Iceland through the Shetland wren, which, 

 although smaller than T. borealis, seems to be larger than 

 T. hirtensis of St. Kilda ; the latter exceeds in size the wrens 

 of Ireland. But no doubt intermediates occur, and a wing 

 received from the lighthouse on the Tuskar rock off the Wexford 

 coast in October, 1888, equals that of a specimen from St. Kilda, 

 whence, perhaps, it may have been a migrant. 



G. E. H. Barrett-Hamilton. 



Kilmanock, Arthurstown, Waterford, Ireland. 



Sexual Dimorphism. 



If Prof. Meldola does not suppose that all spontaneous varia- 

 tions are limited in inheritance to one sex, he is logically bound 

 to admit that the theory of sexual selection does not explain 

 unisexual inheritance. There can be no possibility of verbal 

 juggle in my arguments, because I define a secondary sexual 

 character as one that is affected by castration, one that does not 

 develop normally after removal of the generative organs, a 

 spontaneous variation as one that is not produced by the condi- 

 tions of life. 



But it is not my theory that "the stimulations which pro- 

 duced a male character necessitate the restriction of that 

 character to the male," and therefore I am not affected by the 

 dilemma in which Prof. Meldola thinks I am placed. On p. 94 

 of my book will be found these words : — 



" It is possible that unisexual characters originally developed 

 by special stimulations related to reproduction, tend sooner or 

 later to be inherited in common by all individuals of the species, 

 that, considered in relation to periods of evolution, their sexual 

 limitation is only temporary." 



I fear that Prof. Meldola has not yet sufficiently considered 

 my theory. J, T. Cunningham. 



Penzance, January ii. 



NO. 1630, VOL. 63] 



The theory of sexual selection never pretended to explain 

 unisexual inheritance. Its author started with a fact : — " Inas- 

 much as peculiarities often appear under domestication in one 

 sex and become hereditarily attached to that sex, so no doubt it 

 will be under nature" ("Origin of Species," 6th ed. p. 69).^ 

 Neither does the theory of natural selection pretend to explain 

 ordinary, i.e. bisexual, inheritance. But Mr. Cunningham pre- 

 tends that his theory does explain unisexual inheritance, and 

 having -in spite of the statement contained in the concluding 

 sentence of the above letter — given very full consideration to his 

 views, I have come to the opposite conclusion. I repeat that 

 his theory does not explain unisexual inheritance. 



Mr. Cunningham has now given a further and more restricted 

 " definition " of secondary sexual characters. In this he has not 

 only gone far beyond Darwin, but he has virtually cancelled at 

 least half of his own book. The whole of the evidence that 

 characters developed in one sex are latent in the other was 

 summarised by Darwin in 1868 : — " We thus see that in many, 

 probably in all cases, the secondary characters of each sex lie 

 dormant or latent in the opposite sex, ready to be evolved under 

 peculiar circumstances" (" Variation of Animals and Plants," 

 1st ed. vol. ii. p. 52). All the evidence with regard to second- 

 ary sexual characters which Darwin considered in arriving at the 

 above conclusion was based on cases observed in mammals and 

 birds, and, with his well-known caution, he only admits proba- 

 bility in extending it to all cases. But Mr. Cunningham now 

 has converted Darwin's cautiously expressed probability into a 

 "definition " ! In doing this he has practically wiped out the 

 whole body of material relating to classes other than mammals 

 and birds which he has brought together in his own work. I 

 confess that I have not of late years been able to follow very 

 closely the progress of knowledge in this direction, but, so far as 

 I know, there is no single observation, with the exception, 

 perhaps, of Stylopised bees, which would bring the secondary 

 sexual characters of fishes, reptiles, Crustacea, insects, &c. , within 

 Mr. Cunningham's definition. Is there any known case among 

 these lower groups where the "removal of the generative 

 organs " (to use Mr. Cunningham's own expression) leads to 

 the appearance of the characters of one sex in individuals of the 

 other sex ? 



There is another inexplicable statement in the above letter : 

 " It is not my theory that ' the stimulations which produced a 

 male character necessitate the restriction of that character to 

 the male.' " I must again quote Mr. Cunningham's own remark 

 of December 29, 1900. " My theory is that they (the variations) 

 were so limited in development because they were due to stimu- 

 lations similarly limited " (Nature, January 10, p. 252). If 

 this does not mean that he is attempting to explain the sexual 

 limitation of characters by "stimulations" applied originally to 

 the sex in which they are now developed, then it appears that 

 he has abandoned his fundamental proposition, viz., that his 

 theory explains unisexual inheritance. The restriction of this 

 sexually limited inheritance by considering it temporary instead 

 of permanent, as indicated in the passage quoted by Mr. 

 Cunningham at the conclusion of the above letter, does not 

 affect the argument in any way. We still have to learn how 

 and why the theory of "stimulations" explains unisexual in- 

 heritance, even if the latter be only temporary. 



I venture to think that editorial hospitality has been suffi- 

 ciently taxed in connection with this subject. So far as I am 

 concerned I must beg Mr. Cunningham to consider the discussion 

 as closed. The issue is before the readers of these columns, 

 and I do not think that any further advance is likely to be made 

 by mere iteration and reiteration. I consider that indirectly 

 the author of " Sexual Dimorphism " has done excellent service 

 to the cause of Darwinian evolution by enabling us to realise 

 how a well-conceived and well-worked-out application of 

 Lamarckian principles completely breaks down on critical 

 examination. R. Meldola. 



January 12. 



Very Cold Days. 



The following account of days on which the minimum tem- 

 perature was under 20° (at Greenwich) may be found instructive. 

 There have been 162 of these very cold days in the last sixty 



1 " We may conclude that one cause, though not the sole cause, of char- 

 acters being exclusively inherited by one sex, is their development at a late 

 age " (" Descent of Man," &c. ist ed. vol. i. p. 296). This is the utmost 

 extent of Darwin's application of the evidence in discussing the sexual limi- 

 tation of certain characters. The explanation is based on the hypothesis of 

 " pangenesis " (toe. cit. p. 2S4). 



