THE ARCHOPLASMIC STRUCTURES 235 



limit between sphere and aster, though in other cases it is distinctly 

 present. In the resting cell, on the other hand, the boundary of the 

 sphere is often very sharply marked, so that the sphere appears as a 

 well-defined spherical body. The origin of such a definite sphere from 

 the aster has not been very definitely determined, but Driiner's obser- 

 vations indicate that it arises in the manner described by Van Bene- 

 den, through the disappearance of the more peripheral portions of the 

 astral rays. It is, in other words, the persistent centrosphere.^ 



The genesis of the reticular type of centrosphere is not so well 

 determined. In Xcrcis the aster (maturation-asters, sperm-aster) 

 has at first nothing more than a minute centrosome at its centre. 

 This becomes surrounded at a later period by a large reticulated 

 centrosphere, showing no sign of radial arrangement, that appears 

 to arise by a transformation of the inner ends of the astral rays. 

 A nearly similar account is given by Hill in the case of the sperm- 

 aster in Strongyloccutrotus and Phallnsia. In these latter cases the 

 centrosphere shows no differentiation into cortical and medullary 

 zones. In TJialasseina and Nereis, on the other hand, the minute cen- 

 trosome becomes surrounded by a somewhat vague body distinctly 

 different from the reticulum of the outer centrosphere, and this 

 body perhaps represents a "medullary zone." This body, with the 

 centrosome, corresponds very nearly to the " centrosome " of Ascaris 

 with its " centriole " or central granule as described by Boveri and 

 Brauer ; but in Tlialassema Griffin's observations show conclusively 

 that the minute central granule alone is the centrosome, and that 

 the surrounding body does not persist after division. I cannot 

 avoid the suspicion that the body described by Boveri as the 

 " centrosome " in Echiinis may represent this medullary region of 

 the centrosphere, and that he, like myself, may have overlooked the 

 centrosome. Nor does it seem impossible that the " centriole " or 

 central granule of Ascaris (Boveri, Brauer) may likewise represent 

 the true centrosome. These questions can only be cleared up by 

 further investigation. 



To sum up : The history of the " archoplasmic " structures gives 

 strong ground for the conclusion that attractiou-spheres, asters, and 

 spindle are, like the nucleus, differcntiatioiis of the general cell-netivork, 

 which is, as it zvere, moulded by the centrosome into a specific form. 

 If this be well founded, the word "archoplasm" has no significance 

 save in a topographical or descriptive sense. In this light it is an 

 interesting fact that the aster or attraction-sphere may either persist 

 and divide, like a permanent cell-organ, or may disappear and re-form 

 in successive cell-generations. 



1 The same general result is indicated in the case of plants, though the phenomena have 

 here been less carefully examined. 



