442 



SCIENCE 



[N. S. Vol. XLI. No. 1055 



per cent, of female grandehildren, thus showing 

 that uniform food conditions cause nearly all fe- 

 males to be produced. However, when these roti- 

 fers that were producing almost exclusively fe- 

 male grandchildren on a uniform diet were 

 suddenly put upon a new diet of the green flagel- 

 late, Chlamydomoiias, they almost ceased produc- 

 ing female grandchildren and produced as high 

 as 83 per cent, of male grandchildren. Moreover, 

 if the first few eggs of each female that were laid 

 in the Chlamydomonas diet were discarded all the 

 grandchildren were males, thus showing that a 

 sudden change from a uniform diet to a new diet 

 causes the total suppression of nearly all females 

 and the production of nearly all males. 

 Farthenogenesi^ and Sex in Anthothrips Verhasci: 

 A. Franklin Shull. 



The life cycle of few species of Thysanoptera is 

 definitely and completely known. In general, sex- 

 ual reproduction has been inferred in species hav- 

 ing abundant males, especially if mating has been 

 observed in nature. Such a species is Anthothrips 

 verhasci, the mullein thrips. However, adult fe- 

 males reared in isolation from pupae, and placed 

 on thrips-free plants, have given rise to offspring. 

 These offspring must have been produced parthe- 

 nogenetically. It is not safe, therefore, to infer 

 merely from the abundance of males or the oc- 

 currence of copulation, that any species is sexual. 

 Whether Anthothrips verhasci exhibits both par- 

 thenogeuetic and sexual reproduction has not yet 

 been determined. 



Twenty-eight of the parthenogenetically pro- 

 duced young have reached stages sufEciently ad- 

 vanced to allow their sex to be recognized. All 

 were males. This suggests that the same relation 

 exists between parthenogenesis and sex as in the 

 honey bee and some other Hymenoptera, though 

 other explanations are obviously possible. 

 Sex Control and Knoicn Correlations in Pigeons: 



Oscar Eiddle. 

 Some Internal Factors Influencing Egg Produc- 

 tion in the Shade Island Sed Breeds of Do- 

 mestic Fowl: H. D. GooDALE. 

 Multiple Humarli Births: G. H. Parker. 

 A Note on the Origin of a Color Variety cf Mice: 



Clarence C. Little. 

 A Modification of the Agouti Factor in a Cavy 

 Species Cross: J. A. Detlefsen. (Introduced 

 by W. E. Castle.) 



The agouti character of the wild Brazilian cavy, 

 Cavia rufescens, acts as a single unit in heredity, 

 when transmitted to hybrids between this species 



and the tame species, C. porcellus. This unit char- 

 acter, however, is often modified in the hybrids. 

 The modification is essentially a weakening in the 

 power to restrict black and brown pigments from 

 the sub-apical portion of the hair. The weakened 

 modified agouti character of the hybrids was found 

 to be a recessive in crosses with the normal agouti 

 guinea-pig. The normal agouti of the tame guinea- 

 pig, the modified agouti of the hybrids, and non- 

 agouti, are triple-allelomorphs. 



The Effects of Long-continued Partheno genetic 



Reproduction (127 Generations) upon Daph- 



nids: A. M. Banta. 



The writer has kept pure lines of Daphnia pulex 

 reproducing continuously by parthenogenesis alone 

 for over three years. Some of the lines have now 

 reached the 127th generation. If the sexual cycle 

 is a necessary and essential feature of reproduc- 

 tion in this species the fact should ultimately be- 

 come evident in the reduced vigor in the partheno- 

 genetic lines. In order to discover if any reduc- 

 tion in vigor had actually occurred some "wild" 

 Daphnia pulex were obtained from out-door ponds. 

 These ' ' wild ' ' lines were treated in every way 

 identically (except that no selection was made 

 with them as with the older lines) with the lines 

 already under observation. The age of the 

 mother at the time her first brood appeared, the 

 number of individuals in the fir'st brood and the 

 interval until a second brood was produced were 

 taken as measures of the vigor of the individual. 

 Average values obtained from large numbers of 

 mothers of the ' ' wild ' ' lines and of the selection 

 lines constituted the data finally obtained. 



Measured by these standards, the lines repro- 

 ducing parthenogenetically from the 70th to the 

 92d generation under laboratory conditions pos- 

 sessed somewhat less vigor than wild lines descend- 

 ing from the 1st to the 23d generation under lab- 

 oratory conditions. During the summer (1914) 

 food conditions were quite unfavorable. "Wild" 

 lines descending from the 2d to the 9th genera- 

 tion showed a marked superiority in vigor as com- 

 pared with the lines which during the same de- 

 scended from the 96th to the 103d generation. 

 However on the return of normally favorable food 

 conditions the lines long reproducing partheno- 

 genetically under laboratory conditions actually 

 on each of the three points of comparison appeared 

 to have a superiority of vigor. 



Caswell Grave, 

 Secretary- Treasurer 



[To he continued) 



