GENEBAL CONSIDERATIONS. 313 



cases in which, in consequence of the peculiar conditions of life, th.e linih in 

 question was of little use to the larva at that stage. Parallel cases of this kind 

 of change in the process of development, where there is considerable difterence 

 of shape between the larva and the adult forms, might be cited from other 

 groups of animals. We need here only recall the loss of the larval nervous 

 system and of the integument of the I'ilidiiun, and the origin of these organs 

 in the Xemertine from new rudiments. In other points, regarding Balfour's 

 view, we can only refer our readers to its refutation by P. Mayer (No. 138), 

 which is supported by important arguments (above all, by the reference to the 

 .systematic position and development of Penaeus). 



We tlius, with Claus, regard the Zoaea as a secondarily modified 

 larval form related to peculiar conditions of larval life, which cannot 

 be classed with the series of hypothetical ancestors of the Mala- 

 costraca. 



In the same way the NaupUus also must be regarded as a second- 

 arily modified Crustacean larval form.''" In this case we have to do 

 with a shifting back of specific Crustacean characters into earlier 

 stages. Hatschek (Xo. 15) was the first to point out that, in a 

 derivation of the Crustacea from Pliyllopodadike ancestors, a con- 

 nection of the latter with the Annelida yields the most natural 

 derivation for the whole group. Hatschek supported this view 

 mainly on the agreement found to exist between the body-segmenta- 

 tion and the structure of the adult Crustacean and those of the 

 Annelida, which had already caused Cuvier and Yon Baer to 

 establish the type of the Articulata (Annelida and Arthropoda). 

 Above all, the agreement prevailing in the structure of the central 

 nervous system (segmental chain of ventral ganglia) is so great that 

 we can only refer it to true homology. If, on the other hand, we 

 endeavour to derive the Crustacea through the Nauplius from an 

 unsegmented form of worm, we should be compelled to assume that 

 the points of structure in which tlie Crustacea and the Annelida 

 agree had arisen independently in the two groups (convergence), and 

 thus rested merely on analogy, an assumption which is hardly 

 justified by the facts of comparative anatomy. Besides the agree- 

 ment in structure of the central nervous system, the antennal gland 

 should also be referred to, the homology of which with the segmental 

 organs of the Annelida may be considered as proved. Dohrn 

 (No. 11) arrived at a similar estimate of the value of the Nauplms. 



If, accordingly, we derive the Crustacea (Phyllopoda) from 

 Annelidan ancestors, we must assume for the latter a development 

 through a Twclioplwre stage and through a further Polytrochan 



* [See foot-note, p. 319.] 



