82 BULLETIN OF THE 



cells are derivatives from the yolk-cells, and first appear just before the 

 reversion of the embryo begins. They are abundant along tlie sides of 

 the body, and about the oesaphagus as well as in the dorsal region. 



The yolk during this period is somewhat changed from its early char- 

 acteristics. The corpuscles are undergoing disintegration, and are much 

 vacuolated, which gives them in certain regions a spongy appearance. 

 The nuclei of the yolk-cells, while they have increased in number, are 

 smaller and angular (often triangular) in outline. 



5. The 2^eriod from reversion to hatching. — Few surface changes of 

 importance arc necessary to convert the embryo of the period just de- 

 scribed into the adult. The following are the most obvious : The embryo 

 becomes more closely flexed upon itself (PL II. fig. 11), and the constric- 

 tion which separates the abdomen and the cephalo-thorax is formed. At 

 least two pairs of provisional appendages are modified into as many 

 large spinning mammilla?.* In addition to these two large pairs there 

 is a pair of .smaller median mammilhe, the origin of which I have not 

 traced. The remnant of the tail persists for some time as a post-anal 

 knob ; ui)on the ventral surface appear the infoldings, from wliich are 

 formed the trachefB, and also those of the generative organs ; upon the 

 head the eyes make their appearance. Two or three days before hatch- 

 ing the embryo begins to unroll, and undergoes a moult ; at the time of 

 hatching it is completely straightened. 



I shall now proceed, after this general account of the more important 

 embryonic stages, to the consideration of the development of separate 

 organs and sets of organs. 



III. — Organogeny. 



In the present paper only the following organs will receive attention : 

 (1) the alimentary tract, including stomod;x;um, pharynx, stomach, mid- 

 intestine, stercoral pocket and rectum; (2) the eyes; and (3) the lungs. 



* Balfour ('80, p. 183) lias stated: "The four rudimentary appendages have 

 disappeared, unless, which seems to me in the highest degree improbable, they 

 remain as the spinning mammilla;." Notwithstanding his doubt, I think I have 

 traced the development of two pairs directly into the mammillae. The mammilla?, 

 thorefori', are appendages of abdominal somites, homodynamic with the cephalo- 

 thoracic appendages, and there are consequently six somites condensed into tlie 

 space between the posterior pair of mammillce and the anus. Upon the ventral face 

 the evidences of this are early obliterated, but upon the dorsal surface the poste- 

 rior somites are recognizable by the arrangement of the longitudinal muscles, at 

 least as late as the stage represented in Fig. 70, PI. XL 



