522 A. M. CAlUi SAUNDERS AND ilAKGAKET POOLE. 



occurrence, but it is most interesting to note that Casteel has 

 stated that in Fiona a portion of the left anterior wall of the 

 stomach can be observed to grow more quickly than the 

 corresponding portion in the right. It is not necessary 

 for us to discuss the theories of gastropod torsion, but we 

 may point out that the excess of growth on one side is 

 merely the ontogenetic cause of torsion. What the original 

 phylogenetic cause of torsion may have been we do not know, 

 and possibly never will know. It is not only possible, but 

 it is probable, that the phylogenetic cause was something 

 totally diffei'ent from the actual ontogenetic cause. In tlie 

 more modified members of a group it often happens that 

 certain of the older features in the organisation of the larva 

 get thrown back in development. To some extent this seems 

 to have happened in Aplysia with regard to the torsion, 

 organs seem, that is to say, to develop already twisted. We 

 have indicated the manner in which this occurs with regard 

 to the development of the shell. It must also happen when 

 the visceral loop is developed, for there is no sign of it in the 

 larva when torsion is complete. It must therefore develop 

 already twisted. 



The Coelom. — The mesoderm, as we have seen, appears at 

 lirst in the form of an ill-defined band, but this arrangement 

 is quickly lost. In PI. 22, fig. 3 the cells are becoming 

 irregularly scattered, and a little later are to be found every- 

 where lying between the large yolk-laden endomeres and the 

 ectoderm. In PI. 22, fig. 2 two small mesodermal cells are 

 seen posteriorly, close to the anal cells; these are e and e^ of 

 Carazzi, which, he suggests, may possibly give rise to the 

 rudiment of the genital organ. It is probable that these cells 

 become involved in the formation of the coelom, and that 

 from the wall of the latter at a much later stage the germ- 

 cells arise, but as it is quite impossible to follow these two 

 cells through the larval development, their destiny must 

 remain purely conjectural. As the foot and velum grow out, 

 mesoderm cells pass into them, and there constitute a loose 

 connective tissue; they also form thin epithelial iuvestments 



