vol. i.i Bitter- Davis. — Enteropneusta. 179 



and their nuclei pushed to the inner ends, as shown in PI. XVIII, 

 Pig. 10, or the secretion being discharged, as shown in PI. XIX, 

 Fig. 13. At how early a time, however, in the larval life the 

 secreting begins we do not know, since we have, as already 

 remarked, seen no very young larvae. 



With the beginning of metamorphosis a change sets in in the 

 gastric epithelium, and by the time the first two gill pockets are 

 well developed, but before they have broken through the ecto- 

 derm, this change has come to be profound. Whereas during 

 larval life proper the epithelium was relatively thin, the cells 

 extending through its entire thickness, and the nuclei being 

 approximately all in the same plane, at the stage of metamor- 

 phosis above indicated, the thickness of the wall has very 

 greatly increased, and this increase is accompanied by the irreg- 

 ular scattering of the nuclei through this thickness, PI. XVIII, 

 Fig. 11. It now appears to cursory observation as though the 

 cells do not all extend through the whole thickness, and perhaps 

 they do not. All have, however, become much more elongated 

 than any of them formerly were, and the distribution of the 

 nuclei is chiefly the result of their having taken their places at 

 different heights in the cells. But the most important change 

 is in the structure of the cells. They are now becoming glandular 

 again, but this time the inner ends produce the secretion, the 

 nuclei being pushed towards the outer ends. The final discharge 

 of this secretion into the cavity of the stomach may be readily 

 observed. PI. XVIII, Fig. 12 shows this state of the epithelium. 



The question uow arises, What, exactly, has taken place in 

 this change! Have the new gland cells arisen by transformation 

 of the old ones, or have the old ones broken down and been 

 resorbed, the new ones having had their origin in cells that 

 remained in an indifferent state during the larval secreting 

 period! That the first is the method by which the result has been 

 accomplished is capable of easy demonstration. The outer 

 secreting ends of the larval cells break down, the protoplasmic 

 inner ends become longer, and the nuclei, while moving farther 

 away from the inner ends, do not remain at the very outer limit 

 of the protoplasmic part where formerly situated, but come to 

 occupy a position within the protoplasm. PI. XVIII, Fig. 11 



