156 SCIENCE PROGRESS. 



egg of one species of echlnoderms with the spermatozoids of 

 another species, are a striking case in point. Boveri has 

 recently repeated these observations, with the same resuh 

 as before — namely, that he was able by adding spermatozoa 

 to non-nucleated pbrtions of ripe eggs to raise pure larvae 

 which resembled the 77ia/e parent in every case, while in 

 control series he found that when hybrids were formed with 

 imclcated portions of ova, the resulting larvae exhibited 

 characters intermediate between the two parents. 



As from the standpoint of heredity the nuclear divisions 

 which are associated with the reproductive cells are of 

 special interest, I will ask you to consider a few typical 

 cases. And in the first place, one of the most striking 

 and most uniformly recurring features connected with 

 these cells lies in the numerical reduction to one-half 

 of the previous number of the chromosomes. This reduc- 

 tion may be immediately associated with the differentiation 

 of the sexual cells themselves, as is the case in animals, or 

 it may occur at some other period in the life-history of the 

 orcranism. Thus in all the Arches^oniate and Phanerogamic 

 plants which have been investigated with this object in 

 view, the reduction is manifested in the first division of the 

 spore mother cell ; and if, as in the embryosac (macrospore) 

 formation of Li/itwi, development takes a short cut, the 

 reduced number becomes necessarily shifted to a later 

 period than usual, but it manifests itself at the earliest time 

 after these suppressed divisions would otherwise have 

 occurred. Thus, in Lilmin, the halved number of the chro- 

 mosomes appears in the first division of the embryosac, 

 whilst in Allmm, in which the embryosac mother cell 

 divides once (not twice) before forming the embryosac, the 

 reduced number is apparent at this earlier stage. In the 

 pollen grains, or microspores, the reduction likewise appears 

 in the first division of the pollen mother cells. But it is a 

 noteworthy fact that in Fuctis the reduction takes place at a 

 stage precisely corresponding with that in an animal, and 

 this marks an important difference between Fucus and the 

 Archegoniate plants in which the process has hitherto been 

 most thoroughly investigated. It appears to indicate that 



