326 SCIENCE PROGRESS. 



sion of the extreme difficulty with which they can be made 

 to take a stain at all, and the frequency with which they 

 remain invisible at the spindle apices, although there may 

 be abundant stained microsomes elsewhere in the cell. It is 

 thus evident that certain microsomes stain when the centro- 

 somes do not, or, in other words, that the micro-chemical 

 behaviour is not the same. When, however, the centrosomes 

 do stain, it is equally true that there are usually seen 

 numerous microsometa from which, but for their position, 

 it would be impossible to distinguish them. Consequently, 

 we are at present incompetent to say whether the centro- 

 somes are what they are, by virtue of more than mere 

 position, or not ; because the solution of this difficulty lies 

 beyond our present means of observation, and therefore 

 cannot be determined. The only conclusion warranted by 

 the facts is that the centrosomes do persist through succes- 

 sive generations in many forms of tissue cells. 



On turning now to the few cases in which the spheres 

 have already been found to exist in the Protozoa, it will be 

 seen that the main features of the previous description, 

 respecting the modification undergone by the metazoan 

 spheres, are here repeated with a curious exactitude. In 

 Euglypha alveolata, according to the beautiful figures of 

 Schewiakoff (29), centrosomes appear at the apices of the 

 division spindles, which are destitute of archoplasmic 

 surroundings and at the same time intra-nuclear in position. 

 Beyond the unruptured nuclear membrane, fans of radii 

 spread through the protoplasm and are strictly comparable 

 to those of the metazoan radial envelope. Owing to the 

 intra-nuclear position of the centrosomes, the spindle would 

 appear to originate from the nuclear substance, which 

 probably here replaces in part the archoplasm. 



In the cystonagellate Noctihica there is a large extra- 

 nuclear and typically compound sphere, which is compar- 

 able in every detail, even down to the little light space 

 surrounding the centrosomes, with the compound spheres 

 of the metazoa. When the centrosomes divide and the 

 central spindle is formed, it lies in a groove in the nuclear 

 membrane, exactly similar to that seen in the embryonic 



