D.— ZOOLOGY. 85 



contracts, it is the foot, not the head, which can safely withdraw into it, 

 leaving the head, the most vital part of the body, exposed to attack. 

 This vulnerability of the head in the Nautiloid stage of a Veliger is 

 obviously a defect, but it is not long in being remedied. Head and foot 

 as a whole rotate round through 180° until their relations to the mantle- 

 cavity are exactly reversed. According to Boutan, the whole process of 

 torsion is accomplished in Acmcea in two or three minutes, so that, as 

 Prof. Naef has pointed out, it is difficult to believe that the change 

 is accomplished by ordinary processes of growth alone. A certain 

 amount of true twisting by muscular contractions would seem to be 

 involved. In Trochus (Robert, 1903), the first of all Azygobranchs, the 

 torsion requires six to eight hours. In both forms the shell has already 

 begun its Nautiloid exogastric coil before there is any sign of torsion. In 

 still less primitive forms (e.g. Paludina), as Miss Drummond ' was one of 

 the first to show, the torsion takes longer than in Trochus, and starts at 

 a much earlier embryonic stage, before the shell has begun to coil. It is 

 thus probable, as Prof. Naef maintains, that the slow achievement of the 

 torsion by growth-processes spread over a considerable portion of the 

 ontogeny is a secondary modification. 



The immediate effect of the change, when completed, is to bring the 

 gill-chamber to the front of the larval body, thus enabling the head, with 

 its all-important velum, to be safely withdrawn into it at the first onset 

 of danger. The foot lastly develops an operculum on its hinder surface, 

 which closes the entrance on contraction. 



In the Limpet this rotation is effected during the free larval life, 

 probably as quickly as in Acmcea, its next of kin ; but in Trochus 

 and all subsequent types of Gastropods (Azygobranchia) it takes place 

 in the embryonic phase, so that the Veliger has already undergone 

 torsion before hatching. There can be no two opinions as to the great 

 advance in efficiency shown by the new type of larva as compared with 

 the old. Unfortunately information about the Nautiloid larva of the 

 Limpet and its post-torsional successor is still limited to Patten's observa- 

 tions on specimens reared from artificial impregnations, and neither 

 Patten nor Boutan say much as to the habits of the larvae. It is also 

 difficult to say whether in its retention of a simple prototroch the larva 

 of the Limpet is primitive or secondarily simplified, but the curious 

 changes and variations which have been described in the structure of 

 its prototroch point rather strongly towards the latter conclusion. 

 The larva of Acmcea shows signs of even greater redu tion of its proto- 

 troch, since the ciliary girdle, though composed of two rows of cells, 

 carries only one row of flagella (Boutan). In Fissurella" there can be 

 little doubt on this point, for the larva creeps out of its egg-shell, 

 instead of swimming, and settles down with the least possible delay to its 

 sedentary rock-life, at once proceeding to absorb the prototroch which it 

 has never used in the open sea, and casting the operculum which it has 

 never used at all. The development of Pleurotomaria may some day 



' Paludina, Drummond, Q.J. Micr. Sci., XLVI, 1892 ; Bcutan, I.e., 1899 : Naef, 

 1913, p. 102. 

 " Fisaurella, Boutan, Arch. Zool. Exp., (2) III, 1886. 



