592 WILLIAM PATTEN. 



long after the germinal band is formed, and the appendages 

 have appeared, the yolk itself undergoes a cleavage of greater 

 or less extent. The attempt has been made to show that the 

 cleavage of the yolk is due to some centripetal force of the yolk- 

 cells ; if this was the only cause we should have one of two re- 

 sults dependent upon the relative strength of this centripetal 

 force and the inertia of the yolk particles. If, on the one hand, 

 the force were strong enough to overcome the inertia of the 

 yolk particles, yolk-balls would be formed around each nucleus ; 

 this indeed is often the case, as, for instance, in Lin a and 

 Hydrophilus, and certain Phryganids. On the other hand, if 

 this force were not strong enough to overcome the inertia no 

 cleavage would take place. In other words, provided the force 

 of each nucleus is the same, and they are regularly distributed, 

 either a complete cleavage into a number of segments equal to 

 the number of nuclei will take place, or no cleavage at all will 

 result. For instance, the condition shown in Neophalax would 

 never be realised by such a disposition of forces ; for in this case 

 the yolk is simply divided into two or three segments in which 

 there are a great many nuclei. However, in Mystacides (Zad- 

 dach) there are no more yolk-cells present than in Neophalax, 

 yet there is a complete segmentation of the yolk. There is no 

 doubt that a division of the yolk by increasing the exposed 

 surface would greatly enhance its transformation and absorption 

 by the embryo, and for that reason, other things being equal, 

 we might expect to find a definite relation existing between the 

 extent of segmentation and the duration of the development. 

 This, however, does not appear to be the case, for in fficanthus 

 (one of the Locustina) the yolk is completely segmented, and 

 the eggs which are laid early in the fall do not hatch until the 

 next summer ; while in Blatta ger manic a the yolk is very 

 slightly segmented, and the development lasts but a few weeks. 

 At present, therefore, a great many more facts are necessary 

 before it will be possible to discuss the question with any 

 probability of arriving at a definite conclusion. 



It seems very probable that in all insects at least a large 

 portion of the mesoderm arises by an infolding along the 



