14 H. M. AVOOnCOCK. 



in tlie loos-e connective tissue of the wall. In each case him. 

 is tlie lumen of the diverticulmii, u.v. are the nuclei in the 

 wall, m.f. muscle-fibres, aud c.n. nuclei of the cocloniic 

 epithelium externally. Two much larg-er examples are shown 

 in figs. 18, 19, pi. 2, and it will be seen that the connective 

 tissue of the wall tends to arrange itself in laj'ers around the 

 pai'asite. 



Probable mode of infection. — This situation, together 

 with the non- occurrence in the vascular network, leads me 

 to think that the infection is not by way of the mouth and 

 gut, as in Holothuria, bnt through the cloacal aperture 

 and into the respiratory trees. Probably the nature of the 

 tentacles, which are here of the Dendrochirote type, and the 

 slightly different manner of feeding nccount for this. Whereas 

 a Holotliuria shovels up sand and shell, together Avith any 

 accompanying organic matter, into its mouth, Cucumaria 

 spreads its branched tentacles, like a beautiful net encircling- 

 its anterior end, and waits for living organisms to be entangled 

 in the trap. It then conveys its tentacles with their prey, one 

 after another, to its mouth. Hence the passive spores lying 

 about on the ground are far more likely to be sucked up with 

 the respiratory current through the cloaca and so into the 

 "trees." The fact, too, that the Cncumariee are more 

 sedentary than the Holothurise, not moving about so much 

 in search of food, may explain the comparative scarcity of 

 the parasites. This is, I believe, quite a unique instance 

 of the " casual " or accidental method of infection, which is 

 usually accomplished by the spores being taken in at the 

 mouth when the host is feeding. The excretory acids known 

 to be present in the trees probably perform in this case the 

 function — elsewhere allotted to the gastric or pancreatic 

 juice — of opening the spores. 



llelation of the parasites to the coelomic epi- 

 thelium.— Adults of the typical gregariniform shape never 

 occurred in the trees, but in a very curious position, trying, 

 apparently, to penetrate the coelomic epitlielium, either of the 

 body-wall (most frequently) or of one of the vascular strands 



