]917 Riddle; on Chinook Salmon 323 



The first furrow is perpendicular to the long axis of the disk. It is deeper 

 than the succeeding ones and so can be readily distinguished from them. 

 It is often so distinct in the 32-celled stage as to form a disk shaped like 

 a dumb-bell in surface view {Fig. 9). Three such disks were found among 

 twenty-six 20-hour eggs examined. 



Cavities exist between the cells of the 32-hour disk (Fig. 26). These 

 might represent the segmentation cavity which later exists as the subger- 

 minal cavity. The latter consists of a cavity between the blastoderm and 

 the periblast. The periblast is a layer of granular material upon the 

 surface of the yolk and filled with free nuclei. The subgerminal cavity 

 of the 101-hour egg is shown in Fig. 27. 



FORMATION OF THE BLASTODERM AND PERIBLAST 



The first differentiation of the cells of the blastodisk appears in the 

 formation of the blastoderm. It starts in the 32-hour disk. The superficial 

 cells of the disk at this stage form an unbroken layer and are not markedly 

 different from the cells covered by them. The marginal cells of the 32- 

 hour disk are set above the surrounding germinal material (Fig. 26). 

 The 4i-hour disk contains many cells. The external cells are continuous 

 with one another and are columnar in section. They are different from 

 the other cells of the disk and fuse with the granular material superficial 

 to the yolk. As the development of the blastodisk proceeds, the cells 

 become smaller and finally in the 101 -hour egg (Fig. 27) form a thin 

 membrane which covers a mass of loose cells beneath it. 



The formation of periblast nuclei is characteristic of teleost develop- 

 ment. They were first described by Lereboullet (31) in 185i. Since 

 that time these nuclei have been the object of the study of several au- 

 thorities upon teleostean development. Their research has been directed 

 ylong two channels, their origin and their physiological function. 



In 1876 Klein (28), in a paper on the development of the trout, comes 

 to the conclusion that they arise de novo. 



Hoffman (23) offered an explanation of the origin of the periblast 

 nuclei. He described the first nuclear division as being horizontal and 

 resulting in the formation of two nuclei. The one adjacent to the yolk 

 gave rise to the periblast nuclei and the other to the nuclei of the blasto- 

 disk. Agassiz and Whitman (1) criticized this theory and offered a more 

 satisfactory solution to the problem. They found that nuclei present in 

 the periblast were derived from the marginal cells of the blastodisk, and 

 that they multiplied in the periblast by division. Wilson (49) finds evi- 

 dence supporting this theory in the development of the sea bass. The yolk 

 nuclei found in the Chinook salmon seem to be formed in this manner 

 also. 



