300 J. H. OKTON. 



the sole function of the lophophore is to set up a stream by means of its 

 cilia, and so to bring diatoms and other articles of food to the mouth, 

 and that it has no respiratory function whatever. Indeed, it is difficult 

 to imagine how an interchange of gas could take place through the thick, 

 dense layer of supporting substance." 



There would appear to be little doubt that Shipley is correct, hence in 

 practically all the members of the above-mentioned groups the gill 

 probably functions merely as a water-pump and a food-sieve. In Crepi- 

 dula, most Lamellibranchs and Brachiopods respiration doubtless occurs 

 mainly in the mantle ; while in Amphioxus a gaseous exchange 

 is probably effected chiefly in the coelomic spaces adjacent to the atrium. 

 In Ascidians respiration is doubtless effected in the branchial sac. 



In all these gills there can be no doubt that mucus formation plays a 

 very important part in the process of food-collection. It has been shown 

 in an earlier paper (2) that in Amphioxus and Ascidians the mucus 

 formed in the endostylar groove of these animals is thrown on to and 

 transported along the gill-filaments in thin sheets to serve for entrapping 

 food-particles. In Lamellibranchs and Crepidula it has been suggested 

 that the corresponding mucus formation takes place in the frontal 

 epithelium of the gill-filaments, and indeed swollen cells, which are almost 

 certainly mucus cells, have already been figured in the epithelium of the 

 filaments of Mytilus (see I, Fig. 17, 'passim). Similar globules occur also in 

 the filaments of Glycimeris, Crania (see Fig. 9, p. 299), Terebratula and 

 Rhynchonella (see 6). It is hoped to make this important problem the 

 subject of a special investigation. 



The writer's work, however, on the endostyle of Amphioxus suggested 

 the probable function of a previously enigmatical glandular organ 

 situated along the base of the gill-filaments of Crepidula. An examina- 

 tion of the living animal confirmed the suggestion that this organ is 

 indeed an endostyle, as may be gathered from the description of it in the 

 section on p. 303. 



It is an interesting fact that the spines (clavulse) composing the 

 fascioles of the heart-urchins, Echinocardium and Spatangus, have been 

 found to have the cilia disposed in definite rows along their sides in a 

 manner very similar to that in which the lateral cilia are distributed on 

 the gill-filaments of all the above-mentioned groups. These rows of cilia 

 on the spines do, as a matter of fact, very closely resemble the lateral 

 cilia of gill-filaments in that they lash in a definite direction and with the 

 same wave-like motion characteristic of those current-producing cilia ; 

 they are also situated on opposite sides of the spines, and are indeed 



