324 BULLETIN OF THE BUREAU OP FISHERIES. 



individual variation, but comnjonly elongates transversely to the long axis of the future 

 embryo, endures 4 or 5 days, and then completely disappears. In front of the pit 

 a wide embryonic area is defined by rapid divisions of the surface cells. The latter, 

 which are the direct descendants of the enormous yolk pyramids or hillocks, become 

 distinctly separated into a single stratum of yolk-laden and columnar cells. Below the 

 point of invagination the ingrowing plug of cells expands by rapid divisions of its 

 elements, and like columns of smoke from a steam engine a dense cloud-like mass is 

 spread into the yolk. Many of the cells break loose from the syncytial mass and worm 

 their way through the yolk Uke independently moving amoebae. Many of them degen- 

 erate, while others creep forward and attach themselves to the embryonic area. The 

 cells introduced by invagination gfive rise, in terms of the germ-layer theory, to the 

 hypoblast or endoderm, and to at least a part of the mesoblast. It is almost certain 

 that the yolk-wandering cells receive many recruits from the surface of the embryonic 

 area; the yolk cells iatroduced earlier for the most part degenerate before the stage 

 of invagination is reached. By multiple divisions cell nests are formed, particularly 

 in the embryonic region at the surface, or more commonly just beneath it in the midst 

 of spheroidal masses or balls of yolk. 



Death waits close upon the birth of new cells, and from an early stage to the later 

 egg-nauplius period degeneration is a marked characteristic of this and many other 

 arthropod embryos. Nebulous clouds of chromatin strew the paths of cell migration, 

 and are carried to every part of the egg, where they remain until absorbed. In the 

 early stages at least embryonic layers do not exist, and attempts to reconstruct them 

 out of a mass of rapidly multiplying, degenerating, and moving elements, by the aid 

 of theory and the imagination, have thus far proved neither successful nor profitable. 



The appendages are the first of the distinctly embryonic parts to make their 

 appearance; they are formed by paired tubular folds of the body wall. They pos- 

 sess soUd yolk cores which are gradually absorbed and replaced by mesoblastic cells 

 which migrate from the embryonic region. The hmbs arise in pairs in the following 

 order: (i) First antennae, (2) mandibles, (3) second antennae, (4) first maxillae, and 

 the remaining thoracic appendages in regular succession. The second antenna soon 

 becomes bilobed, the inner branch representing the future long "whip" or flagellum 

 of this limb. The first antennae remain single until shortly before hatching, when 

 the inner flagellum buds out from the inner lower surface of the primary stalk (see 

 p. 226). The optic disks, at first paired rounded areas of rapidly dividing cells, soon 

 become elevated into lobes and form the rudiments of the large eyestalks. The 

 mouth appears at about the ninth day as a median pit on a line drawn through the 

 hinder margins of the buds of the first antennae and before the second antennae are 

 formed. At the tenth day the three pairs of nauplius-appendages are present as 

 buds; a day or two later the upper lip or labrum has grown down over the mouth and 

 a larger fold representing the abdomen and a part of the thorax has grown forward 

 from the region of the thoracic-abdominal plate, marked by the earlier point of invagina- 

 tion. At 14 days of age the latter fold is divided at its extremity, which represents 



