DEVELOPMENT OF EGGS IN MANY-CELLED ANIMALS 337 



tion, which is undoubtedly a process resulting in increased vigour. 

 Presumably individuals developed from unfertilized egg-cells would 

 be less fitted to struggle for existence than those arising in the 

 usual way. The facts ascertained regarding the conjugation of 

 Protozoa throw some light on this (see p. 323). Rare cases are 

 known in which unfertilized ova do develop, and those which are 

 thus distinguished do not as a rule give rise to more than one 

 polar body. It may here be further noted that these bodies very 

 probably have some bearing on the question of heredity, a point 

 which will be dealt with further on. 



Law of Recapitulation. We have seen that certain colonial 

 Animalcules, such as Volvox, help to bridge over the gap be- 

 tween one-celled animals (Protozoa) and many-celled animals 

 (Metazoa), and that in the life-history of one of the latter the 

 egg-cell which constitutes the starting-point is reminiscent of the 

 original one-celled ancestors. But it would be a great mistake to 

 suppose that the Law of Recapitulation enables us to work out in 

 detail the evolutionary history of groups from the stages through 

 which the members of those groups pass in the course of their 

 development. For although, as Milnes Marshall picturesquely put it, 

 an animal may be broadly said to " climb up its own genealogical 

 tree" in the course of its life-history, yet the Law of Recapitulation 

 is by no means the never-failing guide in matters evolutionary that 

 many enthusiastic naturalists once supposed. The expression 

 "once supposed " is used advisedly, for at the present time there 

 is a tendency to discount the law in question to an extent which 

 is hardly warranted by the facts. Balfour, the most gifted em- 

 bryologist of last century, thus expressed the matter (in his Com- 

 parative Embryology) in his usual clear and vigorous way: " Were 

 it indeed the case that each organism contained in its development 

 a full record of its origin, the problem of phylogeny [i.e. evolution 

 of animal groups] would be in a fair way towards solution. As it 

 is, however, the law [of recapitulation] above enunciated is, like all 

 physical laws, the statement of what would occur without inter- 

 fering conditions. Such a state of things is not found in nature, 

 but development as it actually occurs is the resultant of a series of 

 influences of which heredity is only one. As a consequence of this, 

 the embryological record, as it is usually presented to us, is both 

 imperfect and misleading. It may be compared to an ancient 

 manuscript with many of the sheets lost, others displaced, and with 



