THE SPERMATOPHORE IN ARENICOLA 1029 
Objections to Chamberlain’s theory 
The facts adduced in the formation of the spermatophore of 
Arenicola and the evidence cited above to show the independence 
of reduction and the alternation of generations lead me to doubt 
both the validity of his argument and the accuracy of his conclu- 
sion. We are concerned at present only with his argument. 
The fertilized egg of a lily is the first cell of the sporophyte, whether 
it ever divides at all. Consequently, we regard the zygospore of Ulo- 
thrix or Spirogyra and the fertilized egg of Vaucheria or Oedogonium as 
sporophytic structures, even if the first division of the zygote should be 
meiotic, as seems probable. #rom such a simple beginning, we believe 
that the more complex sporophytes with more conspicuous alternation 
have been developed. (Am. Nat. 44: 608). 
LIMITS OF GAMETOPHYTE AND SPOROPHYTE 
The gametophyte and sporophyte generations must be marked 
off, then, it seems to me, by some limits independent of the phe- 
nomena of reduction. A much safer means of definition is to 
return to simple conceptions and designate that generation the 
gametophyte which had its origin in the asexual spore and which 
terminates with the formation of the egg. The sporophyte gen- 
eration begins with the fertilized egg and terminates with the for- 
mation of the asexual spore. Presumably all would agree that 
this means of distinction is the best; the only reason for relying 
on any other criterion is the difficulty of applying this one. 
Yet to insist that the conclusion reached in the study of the 
archegoniates that the 2x number of chromosomes marks the sporo- 
phyte, a law not without exceptions, even in the archegoniates 
themselves, shall apply in the algae, protozoa and higher animals 
too, seems to me unwarranted, for it drives one to the conclusion 
that as great a reduction has occurred in the gametophyte among 
protozoa and algae, Fucus, for instance, as has occurred in the 
whole evolution of the archegoniates from lowly Hepaticae to the 
most specialized angiosperms; and because there is another pos- 
sibility, namely, that the reduction phenomenon, in thecourse of 
evolution, shifts its position with reference to the boundaries of 
