INTRODUCTION. 5 



Examples of such forms have been identified by various embryo- 

 logists in the ovum itself, supposed to represent the unicellular 

 ancestral form of the Metazoa : in the ovum at the close of 

 segmentation regarded as the polycellular Protozoon parent 

 form : in the two-layered gastrula, etc., regarded by Haeckel as 

 the ancestral form of all the Metazoa 1 . 



(2) How far some special embryonic jarval form is con- 

 stantly reproduced in the ontogeny of the members of one or 

 more groups of the animal kingdom ; and how far such larval 

 forms may be interpreted as the ancestral type for those groups. 



As examples of such forms may be cited the six-limbed 

 Nauplius supposed by Fritz Miiller to be the ancestral form 

 of the Crustacea ; the trochosphere larva of Lankester, which he 

 considers to be common to the Mollusca, Vermes, and Echino- 

 dermata : the planula of the Ccelenterata, etc. 



(3) How far such forms agree with living or fossil forms in 

 the adult state ; such an agreement being held to imply that 

 the living or fossil form in question is closely related to the 

 parent stock of the group in which the larval form occurs. It is 

 not easy to cite examples of a very close agreement of this kind 

 between the larval forms of one group and the existing or fossil 

 forms of another. The larvae of some of the Chaetopoda with 

 long] provisional setae resemble fossil Chaetopods. The Rotifers 

 have many points of resemblance to the trochosphere, especially 

 to that form of trochosphere characteristic of the Mollusca. The 

 Turbellarians have some features in common with the Ccelente- 

 rate planula. Some of the Gephyrea in the presence of a 

 praeoral lobe resemble certain trochosphere types. The larva 

 of the Tunicata has the characters of a simple type of the 

 Chordata. 



Within the limits of a single group agreements of this 

 kind are fairly numerous. In the Craniata the tadpole of 

 the Anura has its living representative in the Pisces and perhaps 

 especially in the Myxinoids. The larval forms of the Insecta 

 approach Peripatus. The stalked larva of Comatula is re- 

 produced by the living Pentacrinus and Rhizocrinus etc. 



1 The value of these identifications as well as of those below is discussed in its 

 appropriate place in the body of the work. Their citation here is not to be regarded 

 as necessarily implying my acceptance of them. 



