No. 470] UNITY OF GNATHOSTOME TYPE 



91 



than ever away from an explanation of the origin of the internal 

 sexual organs of the vertebrate body. Certainly, when we have 

 to choose between annelids and amphioxus for an ancestor of 

 the vertebrates, it would be giving up much we have already 

 gained to go back to the vermian type when we have an animal 

 such as amphioxus, possessing many of the vertebrate characters 

 already developed and showing a stage of organization which no 

 one can for a moment doubt is immediately below that of the 

 vertebrates and far removed from that of the annelids and tunicates. 

 It is good occasionally for the zoologist to view^ in the large and 

 in perspective the whole animal and to take note of the interrela- 

 tionship of all its parts, together; in other words, to take a "bird's- 

 eye view" of the form being studied in order that minute and 

 occasional differences, which our incomplete Hnowledge does 

 not yet permit us to explain, shall not be unduly magnified and 

 thereby be given an importance entirely unwarranted, and thus 

 prevent our establishing the homologies and recognizing the real 

 genetic relationships of the form in question. Much that has 

 been said about amphioxus in recent years has been in the nature 

 of zoological quibbling, a playing with non-essentials and an ignor- 

 ing of the fundamental facts of the anatomy and development of 

 this creature. 



The intestinal tract of amphioxus also represents an ancestral 

 condition, which is passed through ontogenetically by higher verte- 

 brates. The liver pouch always arises as an unpaired divertic- 

 ulum of the mesenteron, which later becomes established as a 

 pair of diverticula higher up in the phylum. 



As regards the other features of the intestinal tract, they remain 

 in a very primitive condition, and in the Cyclostonie we have a 

 decided advance towards the condition occurring in higher forms. 

 In the Cyclostome the liver becomes a massive gland, with the 

 characteristic vertebrate structure, but neither amphioxus nor the 

 Cyclostomes possess a pancreas. 



The mouth in amphioxus is extremely ])riniitive and shows no 

 traces of skeletal structures which may yvx he safely hoinologized 

 with the maxillar}^ and mandibular appendages of the Cyclostomes 

 and the vertebrates above them. 



The endostyle, which exists in a high state of development in 

 amphioxus and which is well preserved in the larval Ammocoetes, 



