426 



SCIENCE 



[N. S. Vol. XXV. No. 637 



larval phase, rather than in the late larval or 

 the adult stage, as in the gall-forming hy- 

 menoptera and the plant-lice. It may be that 

 this new method of vegetative propagation of 

 the larvffi of parasitic insects should receive a 

 distinct name. To call it polyembryony is to 

 mislead all who know what is meant by poly- 

 embryony in the original application of this 

 term among plants. 



By recognizing the analogy with the previ- 

 ously known instances of alternation of gen- 

 erations among insects the new discoveries 

 become much more interesting than when in- 

 terpreted as polyembryony. These chains of 

 larvse correspond to the series of so-called 

 parthenogenetic females of the gall-insects and 

 the plant-lice, which are propagated by vege- 

 tative budding. If the fact that all the indi- 

 viduals of the same larva-chain are of the 

 same ses represents a general principle of sex- 

 determination, we must expect to find that 

 this is also true of all the other series pro- 

 duced by vegetative budding. Not only will 

 all the sexual offspring of an individual bud- 

 propagated plant-louse be of one sex, but all 

 the offspring which arise from each original 

 egg of a sexual female. This would mean 

 that what we now call parthenogenetic females 

 are not really females, but represent the two 

 sexes, both temporarily propagated by budding 

 instead of by sexual reproduction. In support 

 of this interpretation we have the fact that 

 even the wingless, larviform, bud-producing 

 plant-lice which are classified as the same 

 species may show two distinct forms, as in the 

 cotton aphis. It now becomes justifiable to 

 suspect that these two forms may represent 

 males and females, and that there may be ex- 

 plained in this way a larval dimorphism which 

 previously appeared altogether mysterious. 

 Adult female plant-lice are wingless, as in all 

 the related families, but the members of these 

 bud-propagating generations are often winged. 

 This has made it necessary to believe that 

 larval females might be winged while the adult 

 females were always wingless, a unique and 

 highly anomalous assumption which the pres- 

 ent considerations may enable us to avoid. 

 It is possible that the bud-producing winged 

 insects may prove, after all, to represent the 



male sex, in spite of their apparent parturi- 

 tion. 



One more confusion of terms remains to be 

 noted. The so-called parthenogenesis of these 

 bud-producing plant-lice is entirely distinct 

 from the parthenogenesis o£ male hymenoptera 

 from unfertilized eggs. If bud-propagation 

 of insects is to be reckoned as parthenogenesis, 

 then the hymenoptera have two forms of par- 

 thenogenesis, one a method of sex-determina- 

 tion, the other a method of vegetative multi- 

 plication. The hemiptera are now known to 

 have a different method of sex-determination, 

 by means of specialized chromosomes and two 

 or more kinds of spermatozoa, so that the 

 plant-lice shoidd not be expected to agree with 

 the parasitic wasps in sex-determination, even 

 though the methods and results of vegetative 

 propagation should prove to be entirely analo- 

 gous in the two groups. 



Mr. W. F. Wight read the last paper, en- 

 titled ' The History of the Cowpea and Its 

 Introduction into America.' This will be 

 published by the U. S. Department of Agri- 

 culture. M. C. Maesh, 



Recording Secretary 



DISCUSSION AND CORRESPONDENCE 



REVERSION INDUCED BY CROSS BREEDING 



Dr. Castle' in an explanation for reversion, 

 thinks that the wild agouti color has been in- 

 troduced through the red parent (when crossed 

 with black) ; and bearing on this is Dr. Daven- 

 port's explanation of reversion to Gallus ban- 

 kiva color, in a cross with white and black 

 poultry, their difference being that Dr. Castle 

 ascribes the reversion of wild color as being 

 added to, and Dr. Davenport as being taken 

 from. 



The latter, in his address at the New York 

 meeting, December 28, explains the phenomena 

 of wild color as being due to the absoi-ption 

 of black by the white, leaving the wild color 

 clean. Perhaps both are right in their re- 

 spective examples. 



We wish to tell of some cases, wherein 

 neither of these explanations can be applied as 

 a cause for the reversions. 



' Science, January 25, p. 151. 



