254 



the iutermediate girdle cell in the a, b and c quadrants and a little 

 later in the d quadrant. At least four rudimentary cells arise from 

 the second quartet, one in each quadrant. All of these cells, when 

 first formed, are small with closely reticulated, deep staining nuclei, 

 and, as they sink in through the body wall and into the cleavage cav- 

 ity, they decrease still further in size and the nucleus is reduced to 

 an intensely staining dot of chromatin. They finally migrate into the 

 endoderm cells, where the cell walls break down, the nuclei disinte- 

 grate and all trace of them is lost (Fig. 3, 1?, C). Although some of 

 these cells, when first formed, are of a fairly good size, in the last 

 stages of their degenerate life it is impossible to distinguish them from 

 polar globules, which in this case suffer a like fate. It is difficult to 

 give an explanation of the behavior of these cells save under the as- 

 sumption that they are vestigial mesenchyme, whose function has been 

 supplanted by large cells formed later in the development. Such an 

 interpretation for rudimentary cells was first suggested by Wilson in 

 reference to the minute cells budded off from the M cells in Aricia 

 and Spio. These, he considers, may be an "ancestral reminiscence" of 

 the time when a much larger portion of entoblast was derived from 

 the cell, as is actually the case in Thalassema or Crepidula. 



While, owing to a lack of necessary material, I have not yet fol- 

 lowed out the later development in full, the distribution of the mesen- 

 chyme in the trochophore of Thalassema seems to be in accordance 

 with Meyer's conclusion, that a good deal of the ectomesoblast per- 

 sists in the adult; for, although a part is differentiated into larval 

 muscles, there are also a great many cells lying on the inside of the 

 body wall and in close association with the mesoblast bands (Fig. 3, D). 

 A careful search of the living trochophore confii-ms Conn's statement 

 that there is no trace of head kidneys in Thalassema. The excretory 

 function seems to be performed by several large floating cells of ecto- 

 mesoblast origin, some of which become attached to the larval muscles. 

 These, in all probability, absorb the metabolic waste, since in older living 

 trochophores they are filled with yellowish brown granules. It is inter- 

 esting, in this connection, to note that DAHLGRtJN has recently dis- 

 covered that the excretory organs of ascidians, in their most primitive 

 condition, consist of similar unmodified mesenchyme, in whose proto- 

 plasm the excretory products form dark granules. 



As has already been indicated in the introduction to this paper, 

 these observations, on the one hand, are of interest when considered in 

 connection with those of Eduard Meyer on the larval stages of anne- 

 lids. According to his observations, the mesoblast, arising from the 



