No. 3.] THALASSEMA AND ZIRPHAEA. 601 



ing agent in mitosis rather than the mere " expression of cell 

 activity." 



As this is an important point, it will be well to treat it a 

 little more in detail. So long as the center of a new astral 

 system coincides with that of a previous one, the mere persist- 

 ence of the focal granules accords equally well with either of the 

 theories representing the centrosome ; for all are agreed that 

 such a structure, once formed, might well persist for some 

 time. A crucial test would seem to be supplied by those cases 

 in which a new system is formed about a point more or less 

 removed from the old center. If the centrosome is but an 

 expression of the forces that give rise to the aster, we should 

 expect to see it deserted in situ at the old center and a new 

 one formed at the focus of the new aster. We have shown 

 that quite the contrary is the case, the new aster starting up 

 about the old centrosome. In this, Tlialassema furnishes the 

 strongest evidence, since the purposeful migration of the cen- 

 trosomes takes place before the formation of the new aster. In 

 thus accepting the physiological theory of the centrosome as an 

 active originating agent in mitosis, we must, however, clearly 

 bear in mind that it by no means necessarily implies the mor- 

 phological theory that it is a permanent cell organ, — though 

 the balance of evidence seems at present to favor the latter 

 view as well as the former. If subsequent research should 

 render the morphological view untenable, this would not, I think, 

 invalidate the physiological theory ; though if, on the other 

 hand, the physiological theory be proved untenable, the main 

 support would thereby be withdrawn from the morphological 

 theory. In other words, if it ever be proved unequivocally 

 that a centrosome can arise de 710VO, it would not thereby 

 follow that the granule thus formed is solely an expression or 

 by-product of the aster formation and not an active agent. 



In all stages previous to the establishment of the equatorial 

 plate, the astral rays can, in favorable preparations, be traced 

 directly to the centrosome, the centrosphere, when present, 

 being constituted by rays that stain more lightly and lack the 

 grossly microsomal structure. From metaphase onward, how- 

 ever, a fine light-staining reticulum becomes developed in the 



