Chromatin as ^'Hereditary Substance^' 3S7 



matid possesses a lash or flagelliiin wliicli arises Croni tlie 

 outermost of a pair of irramiles or ceiitrioles situalcd in the 

 cytoplasm just beneath the surfaee of the cell. In an early 

 stage of the transformation of the salamander sj)erniatid the 

 pair of centrioles moves inward tow;ird the iniclens. the inner 

 member of the pair finally entering the nucleus and becoming 

 the middle piece of the sperm head, while the outer membrane 

 is converted into a ring which finally contributes the undulating 

 membrane of the tail of the completed S])erm. 



Connection between the centrioles and nucleus in the snail's 

 sperm is accom])lished in a different manner. Instead of an 

 inward migration of the ])air as in the salamandt r. the inner 

 member of the ])air sends an ingrowth toward the nuchuis, the 

 centrioles themselves remaining at the surface of the cell and 

 remote from the nucleus. This inij-rowth becomes much elon- 

 gated and produces finally the axial thread of the s))erm, the 

 anterior end of which is embedded in the nuclear part of the 

 head. iNIuch this same sort of thing occurs in many other 

 animals^ both vertebrate and invertebrate. 



There can be no question then that in a large number 

 of animals the centriole of the sperm is primarily quite inde- 

 pendent of the nucleus, and only becomes connected with it 

 as the sperm develops. Consequently, to speculate that orig- 

 inally or ancestrallv the nucleus ffave rise to the centriole 

 and axial thread of the sperm is to go exactly contrary to 

 the most direct and positive evidence we have bearing 

 on the question. To this direct evidence drawn from the 

 study of spermatogenesis, that the centriole is in its origin 

 quite independent of the nucleus, should be added the ex- 

 tensive evidence that the centriole is self-pro))agatlng by 

 division and passes on from cell generation to cell generation 

 somewhat as the nucleus does. But this fact is so familiar 

 a part of elementary cytology as to need no special treat- 

 ment. 



The upshot of this discussion is that while as regards 

 flagella in some protozoa there is solid observational ground 

 on which to rest the theory that chromatic })odles of the 



