26 Papers from the Department of Marine Biology. 



poses, as Smith remarks, "a most peculiar gregarious habit in the 

 larvae," and when we remember that marine larva? are dependent for 

 distribution, so far as we know, on chance vagaries of current rather 

 than an internal migratory instinct, I for one can not suppose that so 

 many larvae should share a common path to their intended host and 

 achieve such remarkable success in fixation. It is true that Delage, 

 experimenting on the development of Sacculina, found that of a brood 

 of Cypris larvae, confined in a small aquarium in company with a crab, 

 large numbers were found attached to the carapace of the crab; but 

 this was certainly due to the stillness of the water in the aquarium and 

 the small space in which the thousands of larvae were confined, condi- 

 tions which would hardly be repeated in nature. The fact that two 

 Sacculinae are occasionally, and even three rarely, found on a single host 

 shows that contemporaneous infection does take place, but it also 

 shows, I think, its comparative infrequency. 



I regret that I can not throw further light on the question of the con- 

 tinuity of the root systems of adjacent parasites, for I think Geoffrey 

 Smith's statement requires confirmation. The time at my disposal 

 was short and material came into my hands only early in my visit, so 

 I did not preserve it carefully for histological study. Nor was I able 

 either to make the search I had intended for the Peltogaster interna 

 stage. But the whole life history is worthy of an exhaustive study by 

 some worker on the Pacific coast having access to this rich source of 

 material. 



THE EVOLUTION OF THOMPSONIA. 



If a budding process is actually proved to exist in Peltogaster and 

 Peltogasterella it helps us to understand how such a form as Thompsonia 

 has risen. When once the continuity of the root system in Thomp- 

 sonia is demonstrated, it becomes impossible to regard it as a primitive 

 representative of the group; it must, on the other hand, be admitted 

 to be the most specialised of all. The root system is undoubtedly a 

 new development unrepresented in the typical Cirripede; but in the 

 course of endoparasitic development in Sacculina it appears before the 

 central tumour, which in its complete development as external sac may 

 still be said to show something of Cirripede structure. But the 

 retardation of the appearance of the external sacs here foreshadowed 

 reaches its climax in Thompsonia. There is here still further exalta- 

 tion of the root system at their expense, for not only do the roots 

 fulfil their nutritive functions, but they also produce germ cells. The 

 root system is, in point of fact, the organism and the external sacs are 

 merely externally placed ovaries, organs of a wholly transient nature. 



It is hardly possible to consider Thompsonia as a direct descendant 

 of Peltogaster socialis, but the former genus may well have had an 



