62 LAMELLIBRANCHIA. 



In the case of Lamellibranchs inhabiting a tube, e.g., the Gastro- 

 chaenidae, this latter is secreted in the same way as the shell ; the 

 free edges of the mantle bend over the wide open shell-valves and are 

 thus able to form the tube (Sluiter, No. 53). The structure of the 

 calcareous tube of Gastrochaena seems closely to resemble that of the 

 shell. 



We still have to mention the relation of the muscles to the shell. The 

 adductor muscles, as is well known, are inserted into the shell and the rela- 

 tion of the epithelium to these muscles is interesting. At the points of inser- 

 tion, the epithelium must either be modified or must altogether degenerate. 

 Ehrenbaum assumes that it degenerates, and ascribes to the muscles them- 

 selves the capacity for secreting the shell-substance (!). As the animal and 

 its shell grow, the adductor muscles, especially the posterior adductor, con- 

 tinually change the position of their attachments on the shell. At their 

 points of insertion, a shell-substance is produced, the so-called transparent 

 substance, and the presence of this substance on the inner side of the shell 

 indicates the course taken by the wandering muscle. But that this substance 

 is secreted by the muscle itself is very improbable, and we therefore prefer to 

 follow the older account of Tullberg, according to which there is, between 

 the muscle and the shell, an epithelium which produces the shell-material. 



B. The Nervous System. 



All the ganglia originate as thickenings of the ectoderm, which 

 subsequently become separated from the latter. The ganglia arise 

 separately and become connected later by commissures. 



It has already been mentioned that the cerebral [cerebro-pleural] 

 ganglion arises in the r rn>rhi>jihor< j larva as a neural plate (Figs. 15, 

 p. 31, 18, p., 36). This consists at first of large closely crowded cells 

 which, by active division, give rise to a multilaminar cell-plate.. The 

 upper layer of this plate which remains continuous with the body 

 epithelium becomes raised up, the lower cell-mass becoming detached 

 from it in the form of two groups of cells. These are the two halves 

 of the cerebral ganglion, the connecting commissures of which no 

 doubt arise in the same way, becoming severed from the ectoderm 

 (this, according to Ziegler, is probably the case in Cycla*). F. 

 Schmidt, indeed, has claimed for the cerebral ganglia of the 

 Uniunidae distinct origins and secondary connection by means of a 

 commissure, a condition which will lie described in connection with 

 the Gastropoda (Chap. XXXI 1.). In Anodonta, the two halves of the 

 ganglion arise near the month and are separated by the stomodaeum, 

 above which the commissure extends as a loop. 



