THE ROLE OF THE ('ELL ORGANS IX HEREDITY 327 



Antcdon (crinoid) the Larvae so produced, contrary to BoverPs results, 

 were maternal in character: they were like the mother, which had pre- 

 sumably contributed cytoplasm only, and not like the father, which had 

 furnished the nucleus. Fertilization by a spermatozoon had here pro- 

 duced a developmental stimulus but no amphimixis (the combining of 

 hereditary lines), so far as could be judged from the appearance of the 

 larvae. Even if the male cytoplasm were admitted to have no heredi- 

 tary role, it nevertheless seemed that the cytoplasm of the egg was clearly 

 so concerned. 



In his work on sea urchin hybrids Baltzer (1910) was able to show 

 why it is that some such larvae are maternal in character while others 

 have the characters of both parents. When an egg of Strongylocentrotus 

 fertilized by a spermatozoon of Sphcerechinus undergoes its first cleavage 

 division the paternal chromosomes behave irregularly; they fail to become 

 incorporated in the daughter nuclei and are lost. Those individuals 

 which develop far enough show maternal skeletal characters. In the 

 reciprocal cross, on the contrary, all of the chromosomes behave normally 

 and the resulting larvae are truly hybrid in character. Thus in the first 

 cross, in which the paternal chromosomes are lost, the spermatozoon 

 furnishes only a developmental stimulus and has no appreciable effect 

 on the character of the new individual; whereas in the second cross, in 

 which the paternal chromosomes are included in the blastomere nuclei, 

 the spermatozoon not only furnishes the developmental stimulus but 

 also contributes paternal characters to the new individual. This is 

 particularly convincing evidence in favor of the view that the chromo- 

 somes are in some way responsible for the development of parental 

 characters in the offspring. 



In a posthumous paper Boveri (1918) reported an additional observa- 

 tion which, he believed, goes far toward explaining the conflicting results 

 of different investigators. He found that egg fragments, and even whole 

 eggs, may often have chromatin in aform that easily escapes observation, 

 but which can exert its usual influence on development. In accordance 

 with his earlier observations, enucleate egg fragments of Sphcerechinus 

 fertilized by spermatozoa of Strongylocentrotus may develop into purely 

 paternal larvae. Most of them, however, are intermediate in charac- 

 ter, resembling the maternal parent also in certain features. Having 

 previously (1895, 1905) demonstrated that the size of the nuclei in 

 merogonic larvae is proportional to the number of chromosomes they 

 contain (see Chapter IV), Boveri was able to show that the nuclei of the 

 intermediate larva 1 are diploid rather than haploid, so that it is clear that 

 the supposedly enucleate fragments in such cases must have contained 

 chromosomes. It is probable, Boveri believed, that the maternal larva 1 

 obtained by Godlewski may be accounted for in a similar fashion. 



It was pointed out by Strasburger (1908), who had come to believe in 



