344 MONCECIOUS MOLLUSCA. 



a varied pattern, painted with delicate tints. In some species 

 from five to sixteen individuals will be united organically 

 in a rosaceous fashion ; other species form a lengthened 

 ribbon, the concatenated individuals being placed side by 

 side, and hence arranged transversely ; others have the ribbon 

 formed of two parallel rows of individuals agglutinated 

 back to back, so that the constituent of one row may dove- 

 tail itself between two of the other ; and others again form 

 chains, in which each member is stretched in the direction 

 of its length, and attached to the one before and the one 

 behind it. The latter are the sea-serpents of the common 

 sailor, for they elicit his notice, as well as the curiosity of 

 the more scientific voyagers. With these compound masses 

 there are found intermingled many single and isolated Salpae, 

 which were long believed to be either distinct species, or 

 the separated individuals of a colony ; for, it was said, that 

 at a certain age, perhaps at full maturity, the chain naturally 

 dissolved, and every link segregated itself.* 



The discovery of M. de Chamisso — a German naturalist — 

 lays open another view of the relationship of the single to 

 the compound organism, and one unseen before in the wide 

 field of physiology. Here the offspring do not resemble the 

 parent at birth, and they remain dissimilar during their 

 whole life, so that the relationship is not apparent until a 

 succeeding generation. The son resembles not the father, 

 but the grandfather ; and in some cases the resemblance re- 

 appears only at the fourth or fifth generation, and even later. 

 This singular mode of propagation has received the name of 

 " alternating generation." It was first observed amongst 

 the Salpae by Chamisso. The Salpae are viviparous, and 

 each species is propagated by an alternate succession of 

 dissimilar generations. One of these generations is repre- 

 sented by solitary or isolated individuals ; the other by 

 aggregated individuals united in groups such as I have 

 described. Each isolated individual engenders a group of 

 associated individuals, and each of these produces in its 

 turn a solitary individual. The solitary individuals are then 

 multiparous, and the associated ones uniparous. This is 

 not the only difference which exists between these two alter- 

 nating generations ; for if we compare the individual in the 

 chain to the naturally isolated one, we find differences in 

 their external form as well as in several peculiarities of their 

 internal organization. A species then has two forms, and 

 you at once perceive how this fact reduces their number in 



* Rang's Manuel, p. 358. 



