ALTERNATION OF GENERATIONS. 537 



man, there is reason to suppose that gemmation can only take place in 

 the earliest periods of existence, perhaps at the epoch of the formation of 

 the mulberry mass. Upon this principle an explanation of the occur- 

 rence of double monsters has been given. 



3D. ALTERNATION OP GENERATIONS. 



It has been already explained that by this phrase is meant that a pa- 

 rent plant or animal will give origin to a form wholly unlike Alternate em 

 itself, and this form, perhaps after the lapse of years, will mation and 

 give origin to another unlike itself, but similar to the original S eneratlon - 

 progenitor. Thus the Salpse present themselves under two different as- 

 pects, the solitary and the aggregated, the latter being produced from the 

 former by being Thudded off in an internal stolon, the individuals being 

 united to one another in an aggregation or chain after they have been 

 separated from the parent. These aggregated salpas alone have sexual 

 organs and produce ova. From each ovum a solitary salpa arises, which 

 repeats the process described again. The solitary salpa, therefore, mul- 

 tiplies by gemmation, the aggregate by generation. Nor is this process 

 confined to animals ; it is also observed in the case of plants. Thus, 

 in ferns, the spore produces the prothallium, which becomes a distinct or- 

 ganism, separated from its parent, and carrying on its nutritive processes 

 independently for itself. From it arises by generation a fern like the 

 original, which, like it, by gemmation, produces prothallia, but never 

 directly produces a fern. Therefore between each fern and its descend- 

 ant a prothallium intervenes, the prothallium arising by gemmation from 

 the fern, and a fern arising by generation from the prothallium. 



After a careful examination of Steenstrup's doctrine of alternations of 

 generation, Dr.. Carpenter concludes that it can not be re- ,, , 



L Explanation 



ceived in the form originally presented, since we should re- of alternations 

 gard a generation as embracing the entire product from gen- of s eneratlon - 

 erative act to act. Indeed, the intermediate forms are often nothing 

 more than sexual organs, furnished or not with an apparatus of locomo- 

 tion, or, in "the more complicated cases, having a mechanism of nutrition 

 attached sufficient for their purpose. The correctness of this interpreta- 

 tion may be illustrated by such cases as the development of medusa buds, 

 which, being first attached to the parent, gradually exhibit the formation 

 of an independent digestive apparatus, and when this has reached a cer- 

 tain degree of perfection, they are separated and swim off, generative or- 

 gans then arising in these buds by which true ova are formed. In the 

 Sentularidas buds are developed in ovarian capsules, and these reproduce 

 in their turn ova by generation. The rate at which gemmation goes on 

 in many of these instances is obviously connected with physical condi- 

 tions, more particularly the degree of temperature and the supply of food. 



