318 The American Naturalist. [April, 
describing the conditions necessary to define the ancestors of the Mam- 
malia, the authors remark: “The mammals have a single temporal 
(zygomatic) arch; the posterior nares are placed far behind, and are 
roofed over by the maxillary and palatine plates; the quadrate is 
completely codssified with the squamosal and quadratojugal ; the occi- 
pital condyle is double, and the entepicondylar foramen is present in 
all the generalized forms. The ancestors of mammals must show the 
same conditions.” It is to be inferred from the context that the 
authors mean that the Reptilian ancestors of the mammals must show 
these conditions. Important exception must be taken to these state- 
ments. The palate is extensively fissured in some Marsupialia, while 
it is closed in the Placodont suborder of the Theromora. The com- 
plete codssification of the quadrate is not to be looked for in a Reptilian 
ancestor, but its reduction must. Such I have shown to be the case in 
the Pelycosauria, in Diopeus and Naosaurus, and Seely has shown it 
to be still more reduced in the Cynodontia. The other characters are 
found in one or another of the Theromora. Hence I believe that the 
opinion that I advanced in 1885, that the Theromora are the ancestors 
of the Mammalia is the correct one. 
Some interesting “asides” are to be found in foot-notes to this paper. 
The authors state correctly that I described two temporal arches in 
Diopeus leptocephalus and, therefore, placed it in the order Rhyncho- 
cephalia, and stated that the Pelycosauria have only one arch, which 
is homologous with the zygomatic arch of mammals. They then add 
“Tt is interesting to note that the latter result was reached by Cope 
(1884) on the identical specimen of (Diopeus) Clepsydrops leptocepha- 
lus.” This statement, ascribing at the very least, gross carelessness to 
the author quoted, is throughout untrue. The ascription of a single 
temporal arch to the Pelycosauria was made by me in the original 
diagnosis of the suborder in 1880 (Proceeds. Amer. Philos. Soc., p. 38) 
four years previous to the discovery of the (C.) D. leptocephalus and in 
the description of the C. natalis, six years previously, in the statement 
“no quadratojugal arch.” This means that the arch present, already 
described by me in Clepsydrops natalis in 1878 (Proceeds. Amer. Philos. 
Soc., p. 509) as a zygomatic arch, was still regarded by me as such. 
In another foot-note the authors make the astonishing assertion that 
what I have called the columella auris in Diopeus is a rib. The skele- 
ton of this specimen possesses ribs of the usual type, however, and 
_neither in this genus nor in any other is there known a rib with a cup- . 
shaped capitulum with a perforation of its peduncle, I have, more- 
over, figured a similar stapes in place in the allied genus Edaphosau- 
rus (Transac. Amer. Philos. Soc., 1892, P1. II, fig. 5a.) with perforation 
