THE FORMATION OF THE ORGANS — THE NERVOUS SYSTEM. 195 



constriction which divides the visceral mass from the cephalic and 

 pedal parts of the body, laterally and somewhat ventrally to the 

 alimentary canal. It is important to note that these two pairs of 

 ganglia, on their first appearance, are quite symmetrical, and that 

 the asymmetry which is so characteristic of the Prosobranchia 

 appears in them only later as a consequence of unequal growth of 

 the different regions of the body. This leads to the right visceral 

 ganglion shifting first dorsally and then over the oesophagus, while 

 the left lies below this tube. In this way, these ganglia become 

 the supra- and sub-intestinal ganglia. 



This process has already been theoretically examined, and illustrated by 

 Fig. 60 A-E, p. 144. Ontogenetically, the process is similar, but is less 

 distinct ou account of the connecting strands of the visceral loop, which are 

 either wanting or difficult to make out. Indeed, the study of the develop- 

 ment of the nervous system is often rendered difficult by the fact that the 

 ectodermal rudiments are so indistinctly marked off from the mesodei-m, as 

 may be gathered both from the text and the figures of older and more recent 

 authors. This difficulty no doubt gave rise to the view that the nervous 

 system in the Gastropoda was of mesodermal origin, which was supported by 

 Bobretzky's observations (No. 11). 



The abdominal ganglion arises, in Paludina and Bythinia, as an 

 unpaired ectodermal thickening on the floor of the mantle-cavity at 

 its posterior end, being found dorsally to the heart. 



As already mentioned, all the ganglia are said to arise independently and 

 to become secondarily connected by commissures. Where, as in Bythinia, 

 the ganglia lie very near each other, they are, according to Erlanger, more 

 distinct in the embryo and only later shift nearer to one another. P. 

 Sarasix, in these very forms, derived the pedal and intestinal ganglia as well 

 as the abdominal ganglion from a common ventral ectodermal thickening, 

 and therefore was able to compare them to the ventral chain of ganglia of 

 the Annelida, whereas the prevailing opinion now is that only the pedal 

 ganglia or rather the pedal strands (which in some Prosobranchia are pro- 

 vided with transverse connecting strands) can be considered as the true 

 homologues of the ventral cords. 



Even in the Pulmonata, in spite of the great concentration of the nervous 

 system peculiar to this division, the ganglia appear as distinct rudiments, and 

 only become connected later. We have repeatedly alluded to the conditions 

 in the Pulmonates, which have recently been very thoroughly examined by 

 A. Henchman and F. Schmidt, although we have dealt principally with the 

 Prosobranchia, in which the processes can be more easily understood on 

 account of the intervals between the ganglia being greater. The formative 

 processes in the Pulmonata agree on the whole with those in the Proso- 

 branchia. 



