April 8, 1904.] 



SCIENCE. 



589 



the yolk of fertilized insect eggs (Doryphora, 

 e. g.) coexisting with healthier vitellophags 

 provided with more rotund nuclei. When we 

 come to the polar body derivatives in his Figs. 

 3, 5, 6, 7 and 8, we recognize the well-known 

 ' dorsal organ,' or remains of the serosa aggre- 

 gating and preparing to pass into the yolk. 

 Between these stages and that of his Fig. 10 

 with nearly completed mesenteron (which he 

 derives from the mesoderm [sic!]) there is 

 another big gap, and so far as the figures go, 

 no demonstrable connection to show that the 

 testes are really derived from such an absurdly 

 improbable source as the ' dorsal organ,' to 

 say nothing of the polar bodies. The only 

 figures in Petrunkewitsch's paper showing un- 

 questionable rudiments of the reproductive 

 organs are Figs. 14, 15, 17, 18, 19 and 20, and 

 in all of these the organs are depicted in the 

 relatively late stages of development that have 

 been figured repeatedly by other authors. 

 Then note the startling migrations described 

 for the drone's testes and their antecedent 

 cells ! The second polar body is at first on 

 the anterior cephalic surface of the egg. 

 The cell derivatives leave this surface and 

 divide into two groups which migrate to the 

 dorsocephalic region and there reunite. The 

 mass thus formed then proceeds caudally 

 along the mid-dorsal line just beneath the 

 blastoderm till it enters the abdominal region, 

 where it breaks up into cells, which migrate 

 ventrally on either side as far as the meso- 

 blastic somites, become entangled with these 

 and are again carried dorsally to the position 

 of the definitive testes. Was ever organ more 

 bedeviled in its development? And what 

 shall we say of the ' critical caution ' not only 

 of taking work of this kind seriously, but of 

 using it for propping up at one of its weak- 

 est points a complicated theory of sex?* And 

 * I allude to that salmagundi (Bull. Mus. Comp. 

 ZooL, Vol. XL., No. 4) in which the disjecta 

 memlra of certain Darwinian, Weismannian and 

 Mendelian theories concerning three such intricate 

 subjects as heredity, sex and parthenogenesis, are 

 stirred to the point of turbidity, garnished with 

 a few accessory hypotheses, and served up in a 

 pamphlet of thirty pages. As if snch messes 

 could be either palatable or digestible! Morgan 

 has presented an excellent criticism of this theory 



if such work on the origin of the drone's testes 

 can be made the basis of a ' Habilitations- 

 schrift,' how implicit should be our faith in 

 the same author's ' Inauguraldissertation ' ? 



Having, as he supposes, established the 

 Dzierzon theory beyond cavil, so far as it deals 

 with the honey-bee. Castle next proceeds to 

 consider the ants, after the fashion of the 

 typical laboratory zoologist whose motto is 

 ' all species look alike to me.' He finds it 

 necessary to admonish me for deeming it ' even 

 a probability ' that workers may develop from 

 the unfertilized eggs of workers. Had he taken 

 the pains to read the observations of Eeichen- 

 bach and Mrs. Comstock with care, or better 

 still, had he acquired a first-hand acquaint- 

 ance with the two insects mentioned by those 

 authors, namely Anergates atratulus and 

 Lasius nigevj he would have seen that his 

 criticism is really as feeble as it is captious. 

 A more careful writer would have observed 

 that Eeichenbach is not a myrmecologist and 

 that his remarks on Anergates, etc., were cited 

 mainly on account of their psychological in- 

 terest as showing the flurry into which a man 

 is thrown on discovering a fact that conflicts 

 with some formidable theory. Anergates 

 atratulus is a rare, monotypic, parasitic ant, 

 which has lost its worker caste and has wing- 

 less, pupa-like males. Obviously, in such a 

 species there can be no nuptial flight, and 

 mating would naturally take place in the 

 nest. Since the worker caste is non-existent, 

 Eeichenbach's, and hence also Castle's, refer- 

 ence to this species, is really irrelevant. In 

 regard to Lasius niger Castle asks : " Is there 

 any reason for supposing that the ants cap- 

 tured [by Mrs. Comstock] had not previously 

 been with males? * * * May we not reason- 

 ably exercise some ' critical caution ' before 

 with Wheeler we conclude it probable 'that 

 worker ants can really produce other workers 

 or even queens parthenogenetically ' ? " It is 

 ('Recent Theories in Regard to the Determina- 

 tion of Sex,' Pop. Sci. Month., December, 1903). 

 It turns out to be merely another case of the 

 old fallacy of juggling the phenomenon to be ex- 

 plained — in this case, sex — back into the germ- 

 cells and then pulling it out again d, la Little 

 Jack Horner, with the naive assurance of having 

 contributed something ' new ' to science. 



