NOTES ON EMBRYOLOGY AND CLASSIFICATION. 439 



We have every reason to hold homoplasy responsible for 

 the agreement (such as it is) between the beak of the turtle 

 and the beak of the bird, between the gill-plate of Lamelli- 

 branchs and the pharynx of low Vertebrates, between the 

 segmentation of some Platyelmia and the segmentation of 

 Appendiculata, possibly for the change of legs into jaws in 

 Crustacea on the one hand and in Tracheata on the other, 

 or possibly for the development of tracheal tubes once in 

 Peripatus, and also a second time in Insects ; but in none 

 of these cases are we led to ascribe anything to homoplasy 

 beyond the production of very general or very simple agree- 

 ments, and until we have reasons for supposing that it can 

 act continuously and cumulatively so as to produce eldboraie 

 correspondences, we are clearly not justified in assigning this 

 power to it in a particular (^se. 



It is chiefly through the valuable pamphlet of Anton 

 Dohrn^ that attention has been directed to the logical 

 necessity of admitting the possibility of very widely prevalent 

 degeneration. At present naturalists are so very generally 

 persuaded (by habit, not by reason) of the universality of 

 progression — that no one has attempted to face the coun- 

 ter theory of universal degeneration which Dohrn put 

 forward. 



Whilst this theory is passed over with silent contempt it 

 appears to me to have as strong a logical position as the 

 theory of universal progression. The evidence of degenera- 

 tion is admitted as conclusive in the case of the parasitic 

 Crustacea and Cirrhipedes. It is equally incontestable in 

 that very large and varied group of non-parasitic organisms, 

 the Tunicata (Urochordate Vertebrata).- The destruction of 

 a very few forms from among the Tunicates would leave us 

 without the evidence of their degeneration. We should then, 

 on the assumption of the prevalence of progression, regard 

 the Tunicates as representing the highest pitch of structural 

 complexity to which their race had attained, and assign 

 them an erroneous position. Obviously, we must be liable 

 to make this mistake with regard to every isolated group, 

 but especially liable to do so in the case of very small groups 

 or isolated genera of simple structure. So strong is the case 

 in favour of degeneration, that at present all that can be said 

 against it and in favour of progression, with regard to any 



' ' Ursprung der Wirbelthieren.' 



* The whole argument as to the Tunicates of course rests on the view- 

 supported by many arguments, that the larval urochord, which many of 

 them possess, is not a larval organ acquired by larval adaptation, but is 

 hereditary and transmitted from adult ancestors. 



