VERMES, CCELENTERA AND PROTOZOA. 169 



pointed out that Balanoglossus possesses, in addition to its 

 vertebrate character of pharyngeal gill slits, a small structure 

 projecting from the dorsal wall of the pharynx into the pro- 

 boscis, which histologically and morphologically resembles 

 the notochord. For this and for other reasons, which it is 

 not necessary to discuss here, he proposed to establish a 

 group called the Hemichordata for the reception of Balan- 

 oglossus, in order to indicate more clearly its supposed 

 chordate affinities. 



A few years later a remarkable colonial animal was 

 discovered among the material brought home by the 

 Challenger expedition, which was at first supposed to be 

 a Polyzoan, but was afterwards proved by Harmer to 

 possess a notochordal structure similar to that of Balan- 

 oglossus and a single pair of gill slits. Cephalodiscus, as 

 this animal was called, was, therefore, transferred to the 

 Hemichordata. 



The genus Rhabdopleura has occupied for many years 

 a very uncertain position in our systems of classification. 

 It forms creeping colonies, which are found attached to 

 ascidian tests and dead corals in deep water in the Nor- 

 wegian and Scotch fjords. Discovered by the Rev. A. M. 

 Norman, in 1868, it has been the subject of memoirs by 

 Profs. Allman, Sars, and Lankester, but it was left 

 for Fowler to discover the existence of a notochord 

 similar to that of Balanoglossus and Cephalodiscus, and 

 to point out other features in which it shows affinities 

 with the Hemichordata. In one important respect, how- 

 ever, it differs from these other two genera, and that is 

 in the absence of gill slits. Whether this is a character, 

 which in itself should be considered sufficient to separate 

 Rhabdopleura from the Hemichordata, is a question which 

 cannot be satisfactorily answered until we have some 

 account of its development, but it is a great gain to 

 know at last in what direction its true affinities lie. 



The two great groups of parasitic platodes — the Tre- 



matoda and Cestoda — -are separated from one another by 



several anatomical characters ; but, of these, the two most 



obvious and important are that, whereas the Trematoda 



12 



