440 



SCIENCE 



[N. S. Vol. XXXIX. No. 1003 



first brood from a young motlier Tvas placed in the 

 experimental tank under carefully controlled con- 

 ditions and while still only a few hours old. The 

 first one of these young to reach the positive end 

 of the tank was selected for the beginning of the 

 + stain and the one last reaching the positive end, 

 or failing to reach it within a given time, for the 

 beginning of the — strain. In a similar manner 

 the selections were made in later generations. To 

 August first for the whole period during which se- 

 lection had been continued the mean reaction time 

 of the + strains of four of the lines was greater 

 than the mean reaction time of the corresponding 

 — strains, this indicating presumably a greater 

 responsiveness to light on the part of the + 

 strains in spite of the selection for the reverse ef- 

 fect. Two of these differences are statistically 

 significant, as they are more than two and one half 

 times the probable error. In fourteen of the eight- 

 een lines the — strains have a higher reaction 

 time than the corresponding + strains and in eight 

 of these the differences are statistically significant. 

 Of the Unes selected for from thirty to fifty gen- 

 erations two have a higher mean reaction time in 

 the -j- strain and nine in the — strain. The two 

 former and six of the latter differences are sta- 

 tistically significant. 



Size Differences in the Spermatozoa from Single 

 Testes: Chables Zelenya and E. O. Faust. 

 Measurements were made of the length of the 

 head in five hundred or more spermatozoa of each 

 of twelve; species of animals. The variation curves 

 plotted from these measurements were used in de- 

 termining the probable presence or absence of size 

 dimorphism. 



In Anasa, Lygceits, Alydus, Musca, Melanoplus 

 femur-ruhrum, Melanoplus differentialis and Phy- 

 tonomus among insects and Pseudomys and Bos 

 among vertebrates the curves are distinctly bi- 

 modal and indicate the presence of two size 

 groups. The inference is made that the group of 

 larger spermatozoa is the one with an X chromo- 

 some and the group of smaller spermatozoa the 

 one without an X chromosome. 



Segregation of Traits in a Pennsylvania Family: 

 WiLHEMiNE E. Key. 



Some Beactions of the Shell of the Pond Snail, 

 Lymncea, to External Conditions: Habold S. 



COLTON. 



A Quantitative Basis of Sex as Indicated hy the 

 Sex Behavior of Doves From a Sex Controlled 

 Series: Oscar EroDLE. 



By his method of controlling sex in pigeons 

 Whitman showed (1) that the first young of the 

 season (spring and early summer) were nearly all 

 males, and yoimg hatched from the later eggs of 

 the season were nearly all females; (2) that if the 

 two sexes arise from the two eggs of any one 

 clutch, that it is in nearly all cases the first egg 

 which produces the male, and the second egg of the 

 clutch that produces a female; (3) that birds kept 

 thus mated and overworked at egg-production tend 

 to produce in succeeding years fewer and fewer 

 males before the appearance of females. 



A study of the sex behavior of the females of 

 one such series (reciprocal cross of T. orientalis X 

 S. Alba) has shown (1) that the females (dark 

 in color) of the alba X orientalis cross are more 

 masculine in their sex behavior (i. e., function more 

 times as males in copulation) than the females 

 (white in color) of the reciprocal cross; (2) tliat 

 females of either cross hatched early in the sea- 

 son, i. e., closest to male-producing conditions, are 

 more masculine in their sex behavior than their 

 own sisters hatched late in the season from eggs 

 produced under strongest female-producing condi- 

 tions; (3) that two full sisters hatched from the 

 two eggs of a single clutch most strongly contrast 

 with each other. The bird from the first or male- 

 producing egg of the clutch usually taking the part 

 of the male to a full 100 per cent. 



The injection (over a period of one month) of 

 extracts and suspensions of ovarian tissue into the 

 more masculine of these females, with simultaneous 

 injections of testicular extract and suspension into 

 the more feminine of the pair, has succeeded in 

 some cases in very strongly reversing the sex be- 

 havior of the pair. The effect persists more than 

 25 days after the last injection. 



The behavior itself, and the effects of the ex- 

 tracts, have been recorded on moving picture 

 films. 



These two results together with our very abun- 

 dant data on the storage metabolism of the ova of 

 these forms, and the initial fact of sex control 

 itself, strongly indicate that the basis of sex is a 

 fluid, reversible process; that the basis of adult 

 sexual difference is a quantitative rather than a 

 qualitative thing. 



Size Inheritance in Babbits: E. C. MacDowell. 



(Introduced by W. E. Castle.) 



To test the hypothesis that the apparent blend- 

 ing inheritance shown by Castle's work on the ear 

 length of rabbits may be interpreted by the mul- 

 tiple factor hypothesis established by Nillson-Elbe, 



