142 bulletin: museum of comparative zoology. 



ture comparable to the radiating system I have described. Huuter 

 sometimes found in ascidians " well developed astral rays " proceeding 

 from the central homogeneous sphere with its centrosomes. Such a cell 

 as that shown in my Figure 3 presents somewhat similar conditions. 



Miss Lewis found in Clymene pi'oducta a centred system, — central 

 granules and fine radiations, — with the additional complication of a 

 sphere having a diameter perhaps one-third that of the cell. The con- 

 densed central region of the cell in my Figure 4 resembles such a 

 sphere. 



The structures most nearly resembling the centred system seen in 

 the earthwoi-m are those described by Biihler ('95) for brain cells of the 

 lizard and ('98) for nerve cells of amphibians and mammals. In the 

 later paper Btihler finds in some cells a centred system with its one 

 or two central granules lying close to the nucleus and fine radia- 

 tions extending toward the cell periphery, and in the same cell he also 

 finds a concentric arrangement about the centre of the cytoplasmic 

 mass, strongly suggesting the sphere of von Lenhossek. In other cells 

 he finds a " spiral figure " similar to that described by Holmgren. These 

 conditions suggest that the spheres of von Lenhossek and Dehler may 

 be structures in no way connected with a centred system of fibrils, 

 and that both structures may be present in the same cells. 



The theory of the structure of the centred system, as proposed by 

 Biihler, is strongly supported by my results. Biihler believes that the 

 centrosome is the insertion point of the stronger fibrils, which may ex- 

 tend to the periphery of the cell. Other fibrils may not insert in the 

 centrosome, but in the granules of the microsome stratum which consti- 

 tutes the boundary of the sphere, or in large granules borne upon the 

 primary radiations. The cell granules are, in general, the insertion 

 points of fibrils, and the granules are larger, the stronger and more 

 numerous the fibrils that insert in them. Accordingly, the centre of 

 the entire system — the centrosome — is the most conspicuous granule. 

 The system of primary, secondary, and tertiary fibrils which I have de- 

 scribed presents exactly these conditions. The fibrils lose in prominence 

 the farther removed they are from the centre. Therefore the primary 

 radiations are oftenest seen. The granules that give rise to secondary 

 radiations may be called secondary centrosomes. We may thus have 

 centrosomes of lower and lower degrees of importance until the limit is 

 reached, — the ordinary microsome. 



Reinkc ('94) finds secondary and tertiary centrosomes. He advances 

 the proposition that a centrosome is potentially present in any micro- 



