652 
PALEONTOLOGY: S. IV. WILLISTON 
Otherwise the skull is very complete and but little distorted. Con- 
nected with the skull was a series of twenty-five presacral, three sacral 
and four caudal vertebrae, all closely united save for a break back of 
the eleventh vertebra, where there was an interval of a few inches, in 
which was found a single centrum and two connected spines, fixing the 
series as not less than twenty-seven presacral vertebrae. An addi- 
tional spine, or twenty-eight in all was collected with the specimen, but 
inasmuch as isolated vertebrae are not rare in the deposit it is possible 
that the twenty-eighth spine does not belong in this series, but its close 
agreement with the adjoining spines makes it probable that it does, 
making twenty-eight presacral in all. In the accompanying restora- 
tion, however, only twenty-seven have been given, the number so 
commonly found in allied genera. 
In collecting the vertebrae lying in the loose shales near the surface of 
the embankment, some fragments were lost connecting the spines, which, 
however, were attached to each other by the thin hard matrix with 
which all the bones were covered. These missing parts, which will doubt- 
less be recovered in the excavated material next season, have been 
represented by oblique lines. 
With this series of vertebrae, extending to the fourth caudal, were 
found parts of the girdles and limb bones which had been more or less 
lost by exposure, or the previous excavations. The preserved parts, 
however, are sufficient to determine their close agreement in size and 
characters with more perfect specimens found isolated near by. Of the 
feet nothing has been determined with certainty in this quarry, but the 
general agreement of the other parts of the skeleton with corresponding 
parts of Dimetrodon — except the vertebrae — make it practically certain 
that these parts are not appreciably different. As in Dimetrodon, the 
tail is, for the most part, unknown, and, as also in Dimetrodon the parts 
actually preserved in position leave little or no doubt of its relative 
shortness. 
The spines of Sphenacodon are of nearly uniform length to the base 
of the tail. Those of the neck increase a little, both in thickness and 
length to the beginning of the dorsal region above the scapula, where 
they are the longest and stoutest in the whole series. They decrease 
very sKghtly in length posteriorly and grow thinner and less stout. 
With the first presacral they become markedly thinner and narrower 
and with the caudals grow rapidly more slender and shorter, leaving lit- 
tle doubt of the weakness and shortness of this part of the body. The 
vertebrae here also are shorter and smaller than those in front of them. 
The ends of the cervical ribs have been lost, but with those exceptions 
