Mr. F. S. Conant on the Ghcetognaihs. 209 



structures. A cross section in tliis case would show a large 

 tube — the oesophagus — enclosing two smaller yet relatively 

 large tubes — the diverticula. The latter bear on the surface 

 that is turned toward the digestive tract an epithelium like 

 that of the rest of the tract, and they thus represent exactly 

 the structure that might be produced by pushing external 

 diverticula back upon themselves into the digestive tract. 

 Much careful examination failed to reveal any specific differ- 

 ences between the animals having external and those having 

 internal diverticula. It may be that observation of living 

 specimens, however, would show that we have here two 

 species. There is a possibility also that the two forms are 

 different phases of the same structure in one and the same 

 animal ; but this seems to me improbable. Some specimens 

 having internal diverticula show a formation at the base of 

 these suggestive of external diverticula as well. 



The function of such internal diverticula can hardly be 

 surmised. The lumen, which is connected with the body- 

 cavity, is often found to contain coagulated coelomic fluid. 

 What strikes one most prominently, however, is that they 

 would seem of necessity to be very much in the way in the 

 act of swallowing. 



III. Classification. 



In May 1895 E. B6'aneck published in the * Revue Suisse ' 

 (tome iii. fasc. 1) an article entitled " Les Ch^tognaths de la 

 Bale d'Amboine," in which he devoted several pages to a 

 critical review of the systems of classification. After com- 

 paring the defects in the classifications of Langerhans (fol- 

 lowed by Strodtraann* in 1892), Hertvvig, and Grassi, he 

 came to the conclusion that the first offers the most advan- 

 tages in our present state of knowledge. 



At about the same time (' University Circular,' June 1895) 

 I was having difficulty in finding a place for Spadella schizo- 

 ptera in any of the systems. It seemed to me that Grassi's 

 classification met the case the most satisfactorily, except that 

 he unfortunately interchanged the names of his two genera, so 

 that his Spadella very nearly represented the Sagitta of other 

 authors. 



Now, however, having seen B^raneck's paper and having 

 studied more forms, I am inclined to admit the full weight of 

 the arguments against Grassi's system. Dividing a group of 

 animals into two genera, one characterized by certain features 



* Archiv fiir Naturgesch. Iviii. (1892). 



