306 J. H. ORTON. 



be seen that in both there are two rows of similar long glandular cells 

 with basal nuclei, separated by a row of ciliated cells wnth apical nuclei. 

 The function of the intermediate row of ciliated cells in Ampliioxus has 

 been shown to be that of passing on to the gill the mucus formed by the 

 gland cells (see 2, Fig. 3, and p. 25), and the function of the corre- 

 spondmg cells in the endostyle of Crepidula is the same. 



It has now been shown (1) that the endostyles in Amphioxus, 

 Ascidians, and Crepidula, and its allies * show a remarkable resem- 

 blance, and also that the ciliation and internal supports of the 

 gill-filaments in Gastropods, Lamellibranchs, Amphioxus. Ascidians, 

 and Brachiopods are essentially the same. Consequently it is clear 

 that the homological "f" value of any of those characters cannot 

 be great since they are present in widely divergent groups, and indeed 

 would appear to be merely similar adaptations to a similar method 

 of feeding. Therefore it follows that organs present in dii^erent animals, 

 but having the same functions, are never necessarily homologous, since 

 practicallv identical organs are here shown to occur with the same 

 function in groups which are undoubtedly only remotely related. 

 These groups are indeed so remotely related that the characters of 

 the ciliation and internal supports of the gill-filaments, on the one 

 hand, and the endostyle in Crepidula and in the above-mentioned 

 Protochordates.J on the other hand, must have arisen independently 

 in each case to meet similar requirements. Such plasticity in 

 organisms is indeed remarkable, and its demonstration necessarilv 

 increases the difficulties in the already difficult problem of determining 

 what organs in different animals really are genetically related. For 

 instance, the presence of an endostyle in Crepidula confounds all the 

 arguments that we can advance at present in support of the theory 

 that the parabranchial ridges of the Enteroprensta are homologous with 

 the endostyle of Amphioxus (see Willey, 15). These organs may possibly 

 be homologous, but we have no means of determining with any certainty 

 that they really are : and a similar uncertainty must exist in other 

 similar cases. 



* It is not improbable that an endostyle may be present also in many other aquatic 

 Oastropods. 



t The term Homology is used with the meaning Lankester gives to the term Homo- 



geny (see 16). 



t It is not contended that the endostyle has arisen independently in Amphioxus and 

 Ascidians, for the well-known reason that development in the latter group indicates a not 

 extremely remote relationship of that group with Amphioxus. 



