KOFOID AND SWEZY: UNARMORED DINOPLAGELLATA 61 



A homology between the nematocysts of Polykrikos and Nematodinium and 

 the trichocysts of the ciliates appears also to be untenable, since the structure 

 of the two organelles differs so greatly. The origin of the latter structures is 

 still obscure, Maupas (1883) and Sehewiakoff (1889) claiming an ectoplasmic 

 origin for them, while Brodsky (1909) found their origin in the endoplasm. 

 Their position is consistently in or closely abutting upon the ectoplasmic layer 

 in most if not all ciliates. In Polykrikos, as in the genus Nematodinium , the 

 position of these organelles is not definitely localized. Neither can any definite 

 statement be made as to whether their origin is ectoplasmic or endoplasmic. 

 In the undischarged trichocysts of Frontonia leucas (fig. H, 2) a certain super- 

 ficial resemblance to the nematocysts of Polykrikos is found. This resemblance 

 is entirely lost when the trichocyst is discharged (fig. H, 1). A long filament 

 is throw^n out beyond the surface of the body, while the posterior portion of the 

 trichocyst also elongates, the whole structure becoming several times its original 

 length, diminishing in width at the same time. 



In the Cnidosporidia the polar capsules, while resembling the nematocysts 

 of the dinoflagellates in general appearance, yet differ from them in their fixed 

 position, and in the fact that they are formed within a capsulogenous cell, from a 

 definite part of the developing sporocyst. In the spores of Myxoholus (fig. H, 5) 

 the coiled thread of the nematocyst, lacking the differentiated introvert of those 

 in the dinoflagellates, is everted as a slender, tul)ular extension of the capsule. 

 This may be thrown completely out of the spore at the time of discharge 

 (Keysselitz, 1908). 



In the nematocysts of the Coelenterata the anatomical features show only 

 a slight advance over those found in the dinoflagellates. The introvert may 

 be armed with spines variously arranged on the coiled filament when it is 

 everted (Toppe, 1910). In the resting stage the appearance of the whole 

 structure is not unlike that of the nematocysts of Polykrikos (fig. H, 3). The 

 accessory structures with which they are associated in the coelenterates, as the 

 cnidoblast, form one of the chief differences between the two organs. The 

 question of homology l)etween the two structures is one that needs more critical 

 data for its solution. The lack of a definite cnidoblast would not necessarily 

 point to an absence of homology, since that would probably be a secondary 

 acquisition, resulting from the multiplication of cells of the body and their 

 specialization, and unnecessary where one cell carries on all the functions of 

 the body. A certain amount of this specialization has taken place in the 

 Cnidosporidia, where the body of the developing somatella becomes divided 

 into two parts, each with a definite number of nuclei with definite functions 

 in the developing spore. Here also a further advance in the development 

 of the nematocysts has taken place in that they have acquired an enveloping 

 capsulogenous cell. 



