246 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 



stages of disintegration ; from the character of the epitlielial lining of 

 the chamber, sections of which prove it to be composed, in part at least, 

 of more or less columnar cells ; and, finally, from the wide communica- 

 tion of this cavity with the true stomach, this communication being in 

 no clear way marked off from the two connected cavities, I judge that 

 to a considerable extent the branchial membrane has become digestive. 



The branchial chamber of the ascidian being, as is now universally 

 recognized, the highly modified anterior end of the digestive tract, if the 

 class be supposed to have had an ancestor in which the region was 

 truly digestive ; and if the conjecture that in Odacnemus the branchial 

 membrane has secondarily acquired the digestive function, we should 

 have here the very unusual instance of an organ resuming its original func- 

 tion after having become highly modified for a wholly different function. 

 The data are rather too dubious to make profitable much speculation as 

 to whether this resumption of the original function could be attributed 

 in any wise to a true reversion ; that is, to the influence of a long-dor- 

 mant character. But assuming such a resumption to have taken place, 

 the fact might be more natm-ally accounted for by the influences, direct 

 or indirect, of life at the great depth in which the animal lives. It 

 seems that for numerous deep-sea ascidians, respiration does not demand 

 the service of any such elaborate mechanism as that possessed by the 

 typical shoal-water members of the class. In a considerable series of 

 species, widely separated taxonomically, the branchial membrane is much 

 reduced in one w-ay and another. Instances of tliis are furnished by 

 Ascopera, Corynascidia, Hypobytkius, and a new and remarkable form 

 found off the coast of California at 2,000 fathoms, which I have studied 

 but have not yet described. Wherefore this diminution of importance 

 of the branchial organ for respiratory purposes, is not obvious ; given 

 the fact however, there would appear no special difliicnlty in conceiving 

 that the cavity might gradually be turned over to the food-taking and 

 digestive functions. 



Continuing our examination of the viscera, we find a wide but very 

 short passage from the posterior end of the branchial cavity (Fig. 4, oe., 

 Plate 2) into another still more capacious chamber lying immediately 

 beneath the floor of the cavity already described. The passage-way is 

 clearly the oesophagus, and the large chamber the true stomach ; or 

 more exactly a stomach-intestine ; for it is not sharply set off from a 

 true intestine. This chamber (st.-in., Fig. 5, PI. 2) extends forward, 

 narrowing down rapidly to a very small, short rectum. The exact posi- 

 tion of the anus I have unfortunately not been able to find ; though it 



