66 



SCIENCE 



[N. S. Vol. XXIX. No. 732 



In what measure such a formulation of 

 the facts may be adequate will find its test 

 in the facts of parthenogenesis ; and while 

 these are not sufficiently known to give a 

 positive result, they seem in the case of 

 animals to be, on the whole, not out of 

 harmony with it. "We must clearly dis- 

 tinguish between parthenogenesis with and 

 without reduction, for in the former ease 

 one X-element is eliminated, while in the 

 latter ease both are presumably still pres- 

 ent. Parthenogenesis preceded by the 

 formation of a single polar body without 

 reduction occurs in the summer genera- 

 tions of aphids, phylloxerans, daphnids 

 and rotifers, and in all of these females 

 are produced, since the female chromo- 

 some-combination persists unaltered. The 

 male-producing eggs likewise form but one 

 polar body and do not undergo a general 

 reduction. As already stated, however, in 

 the aphid or phylloxeran they eliminate 

 one chromosome (the X-element) and thus 

 produce the male combination. 



The crucial test of the assumption lies in 

 the parthenogenesis of eggs which form 

 both polar bodies; for if it be correct the 

 egg which develops with the reduced or 

 haploid number of chromosomes should 

 produce a male, and that which develops 

 with the diploid number a female.^^ The 

 facts are not yet known with sufficient ac- 

 curacy to admit of a decision, but with 

 one or two possible exceptions the best 

 known cases seem to be, on the whole, in 

 harmony with this. In Bhodites the eggs 

 are usually female-producing, and were 

 long since described by Henking as under- 

 going a preliminary coupling of the 

 chromosomes ; but the diploid number is re- 

 stored by a doubling of the chromosomes 



" There is, however, a possibility that in female- 

 producing eggs reduction might occur in respect 

 to all the chromosomes except the X-pair, which 

 would form the converse case to that observed by 

 Morgan in Phylloxera. 



previous to cleavage. Henking interpreted 

 both divisions as equational, and assumed 

 that no qualitative reduction occurs. More 

 recently Doncaster describes the female^ 

 producing eggs of the saw-fly PoecUosoma 

 luteum as also developing with the diploid 

 number, in this case without a previous 

 coupling and doubling. Both these cases 

 are therefore in harmony with the assump- 

 tion. In the ant and bee the male produ- 

 cing eggs were supposed by Henking 

 (Lasius) and Petrunkewitsch (Apis) to 

 undergo reduction followed by doubling, 

 as in Bhodites, which would be a contradic- 

 tion to the assumption ; but neither of these 

 conclusions is borne out by more recent 

 work. Schleip's studies on the ant 

 (Formica) leave little doubt that the un- 

 fertilized eggs of the workers develop, at 

 least up to a late stage, with the reduced 

 number (24) and that the fertilized 

 female-producing eggs of the queen de- 

 velop with twice this number. In case of 

 the bee, likewise, the work of Meves on the 

 spermatogenesis renders it almost certain 

 that Petrunkewitsch was misled, the num- 

 ber 16, which he observed in the cleavage 

 of the drone eggs, being the reduced num- 

 ber. The ant and bee therefore also fall 

 into line with the assumption. A diffi- 

 culty, on the other hand, appears in Don- 

 caster's results on the parthenogenetie eggs 

 of a saw-fly (Nematus rihesii), which is 

 said to produce usually males, but some- 

 times females. Doncaster makes the ex- 

 tremely interesting observations that there 

 are here two types of maturation, both 

 polar bodies being formed in each case, but 

 in one type reduction occurs, in the other 

 it does not. If we could assume that the 

 former type is male-producing, the latter 

 female-producing, the general assumption- 

 would receive a strong confirmation; but 

 the spermatogonia are described as divid- 

 ing with the diploid number. If this is 



