142 A. P. MATHEWS. 



treatment do not s\vell and disappear as they ordinarily do on 

 crushing the egg, but persist as granules and may be separated 

 from their surrounding matrix. They then acquire a very active 

 Brownian movement and float off as granules. 



That they exist as granules in the egg is shown also by their 

 behavior during mitosis and in the centrifuge. In Arbacia, the 

 granules are some of them pigmented. During mitosis, these 

 granules do not remain evenly distributed over the surface 

 of the egg, but collect along the furrow of segmentation. Also 

 when the astral centers are formed the granules either move 

 away from the neighborhood of the asters, as has been described 

 for other eggs, or else they are dissolved in these areas. At any 

 rate near the asters and spindles only the clear protoplasm with- 

 out granules is to be found. 



Furthermore, when the large astral radiations extending to 

 the surface of the egg are formed, it may be seen that these 

 radiations consist of the clear homogeneous matrix. The gran- 

 ules are arranged in rows between the broad clear rays. When 

 oxygen is taken away the granules move in again where the rays 

 were. 



Further evidence is obtained if the eggs are centrifugalized in 

 the method discovered by Lyon at Woods Holl, this summer. 

 By centrifugalization the granules are separated into zones of 

 granules, as has been described by Lyon and Lillie and the clear 

 protoplasm forms one zone by itself. 



The clear protoplasm which forms the matrix in which the 

 granules are, is oftentimes extremely viscid. This is shown in 

 several ways. The granules possess, as long as the egg remains 

 normal, no Brownian movement, showing that they must be in 

 a viscid matrix. It is only when the egg is partially liquefied 

 that they acquire Brownian movement. The viscidity is also 

 shown by observations on living star-fish eggs partially deprived 

 of oxygen. Long viscid strands can then be seen extending be- 

 tween the blastomeres, with here and there granules embedded 

 in them. 1 If these eggs are beaten with a rod, it is frequently 

 possible to draw the protoplasm out in long clear threads. 



1 See Mrs. Andrews : Journal of Morphology, 1896, xii, p. 367. 



