180 University of California Publications. [zoology. 



shows a stage in which the larval glandular condition has nearly 

 disappeared, while the cells have become in a corresponding 

 measure protoplasmic. The protoplasmic part stains distinctly 

 though not intensely, while the outer glandular ends, o. c, 

 remain wholly unaffected by the stain, and have the peculiar 

 refractive properties characteristic of lifeless, degenerating 

 cell substance. At a stage a little later than that shown 

 in this figure, the last trace of the larval glandular condition has 

 wholly disappeared, though the cells have not yet become 

 secretory at their inner ends. Then, finally, by the time the first 

 pair of gill pockets has broken through the ectoderm, the cells 

 have become glandular at their inner ends, and have begun to 

 discharge their secretion into the cavity of the stomach. An 

 early stage of this condition is shown by PI. XVIII, Fig. 12, 

 already referred to. 



Whether this interesting two-fold function of the gastric 

 epithelium at different periods in the life of the animal holds 

 good for all species in which the tornaria occurs is not certain, 

 though in all probability it does. Hpengel has given us enough on 

 the species studied by him to show that at least the earlier larval 

 condition is there the same as here. His remark, however, 

 which concludes what he has to say on the migration of the 

 nuclei toward the inner ends of the cells, namely, that they 

 "maintain this position in later stages," Spengel, '93, p. 399 

 seems to indicate that he did not follow the full course of the 

 functional and structural changes of the epithelium. Morgan, 

 *93, describes and figures the gastric epithelium in the species 

 studied by him as composed of tall columnar cells, and in his sec- 

 ond paper, Morgan, '94, wherein the metamorphosis is treated 

 in detail, he makes no special point of the change of this epithe- 

 lium. In this last paper, however, he gives us, under the heading 

 Growth, a discussion at some length of a number of interesting 

 developmental questions presented by the career of the tornaria as 

 a whole. He here assumes that the "gelatinous fluid," as he calls 

 it, that occupies the blastocoel increases in quantity during 

 larval life, and through secretion from cells somewhere. 

 "Whether, however," he says, "these cells are ectoderm, mesen- 

 chyme, or end.od.erm, I do not know." 



