240 GILBRET 0. r.OURNR. 



Ascidians. The amoebocytes of the test nre generally elongate- 

 oval in form, sometimes irregular and amosboid in shape; 

 each contains a small, deeply staining nucleus, situated to one 

 side of the cell; the cytoplasm is vacuolated, and in most 

 cases contains granules of various forms. In some cases the 

 granules are of a bright yellow colour, and these will be 

 described more particularly further on. The hair-like pro- 

 cesses or villi of the test do not differ from those described in 

 many other Molgulid?e and in some Cynthiadte. The villi 

 are longest and most numerous on the rounded posterior 

 (aboral) end of the body ; they are almost equally abundant, 

 but shorter and thicker, on the outer faces of the arms ; they 

 are short and more sparsely distributed over the rest of the 

 body. The villi are hollow, and most of them contain 

 amcfibocytes ; some of them, especially the longer ones at the 

 posterior end of the body, end in enlargements, to which 

 grains of sand are fixed, like those of Molgula (Anurella) 

 roscovita figured by de Lacaze Duthiers. The villi of the 

 arms and body, however, have bluntly pointed or rounded 

 extremities, and sand-grains, sponge spicules, etc., are attached 

 to their sides by a flocculent yellow material, which, as I 

 shall show presently, is probably excreted by some of the 

 amo3bocytes of the test. The substance of the test is pene- 

 trated in all parts of the body by somewhat sparsely dis- 

 tributed, knobbed, and tassel-like vascular processes of the 

 body-wall. These will be more conveniently described in 

 connection with the ectoderm, but it may be mentioned that 

 they do not penetrate, or, as far as I could determine, enter 

 into any relations with the hair-like villi of the test. 



The Ectoderm. 



The relations of the external ectodermal epithelium to 

 the test could not be very satisfactorily studied owing 

 to the shrinkage of the tissues of the body-wall and their 

 separation from the test. In some places the epithelium 

 was adherent to the test; in others it remained adherent 



