930 The American Naturalist. [November, 
have lost segments offers fewer difficulties than the assumption 
that they acquired segments to develop into the vertebrate 
ty pe. 
If morphological evidence has any value whatever, the con- 
clusion seems to me inevitable that the Tunicates and Amphi- 
oxus constitute a natural group, Atriozoa, which are somewhat 
distantly allied to the Vertebrata. 
2. THE CHORDATA. 
Zoologists are generally agreed that the Chordata is a nat- 
ural group, in the sense that the Tunicates, Amphioxus and 
Vertebrates are more closely allied to one another than with 
any other known animals. I can find no reason for dissent 
from this view, it therefore suffices to follow those who have 
insisted that the problem of the origin of Vertebrates is part 
of the problem of the origin of Tunicates. 
5. THE PRINCIPAL THEORIES OF THE ORIGIN OF VERTEBRATES. 
It is, in my judgment, unnecessary to discuss the theories of 
Adam Sedgwick as to the origin of metameric segmentation,” 
or of Hubrecht that the Nemertinesare the ancestral type from 
which Vertebrates have sprung. The latter I have discussed 
previously. Of the former I will only remark in passing that 
it seems strange that Sedgwick has persistently denied the 
concrescence theory of the embryo, and yet puts it forward as 
a new idea upon which he founds his hypothesis of metamer- 
ism. 
There are five theories known to me which certainly require 
` consideration, namely, Gaskell’s, Patten’s, Bateson’s, and the 
Appendicularia and Annelid theories. 
A. Gaskell’s theory was ably presented in Dr. Gaskell’s ad- 
dress, delivered last year at Liverpool before the Sections of 
Physiology and Zoology. The main points in this theory are 
that the Vertebrata arose from Crustacea like animals, the 1n- 
testine of the latter becoming the tubular nervous system of 
* Of course, not because the theory does not need consideration, but because > 
deals not directly with the origin of Vertebrates, but with the origin of segmented 
animals. 
