19 
The presence of thirty-two presacrals in the Edmonton specimen appears 
at once to indicate its generic distinctness. Unfortunately the skeletons 
upon which Marsh based his original determinations no longer exhibit a 
verification of thirty as being the correct number. From other specimens now 
known from the Lance formation it would seem that perhaps Marsh was 
mistaken in his original count. The type of T. annectens (Marsh) now 
exhibited as a panel mount in the United States National Museum, No. 
2414, has thirty presacral vertebrae, but four of these, the sixth and seventh 
cervicals and the seventh and fourteenth dorsals, are entirely restored, 
and this fact raises grave doubt of the column being complete. The 
companion specimen, No. 2182, belonging to the Yale Peabody Museum, 
is packed away and unavailable at this time. A study of the illustrations 
of the mounted skeleton published 1 by Beecher clearly shows nineteen 
dorsals, and certainly more than eleven cervicals. The neck bones are 
somewhat disarranged in this mount and it is, therefore, difficult to determine 
the precise number, though there appear to be twelve if not thirteen. 
Both of these specimens have nine vertebrae in the complete sacrum. 
The two mounted skeletons in the American Museum of Natural 
History, New York, identified as Diclonius mirahilis Cope and also from 
the Lance formation, are likewise deficient in positive information as to 
the complete presacral formula. Under date of March 15, 1923, Matthew 
replied to my inquiry as follows: “The cervical-dorsolumbar formula on 
our two mounted skeletons is 16-18 in each; but the number of dorsals 
appears to be uncertain in the Cope skeleton (No. 5730) and the number 
of cervicals in the Sensiba skeleton (No. 5894), as two or three are restored 
in each case. There is a third skeleton, No. 5058, which has either 34 or 
35 presacrals, as far as I can make out in its present condition.” 
From the above review of the known Hadrosaurian specimens from 
the Lance formation, with which the Edmonton specimen appears to have 
its nearest affinities, we have on the one hand one species with two less 
presacral vertebrae, and on the other hand a second species which has two 
more. If these formulae are found to be correct they represent differences 
of more than specific importance. 
The cervical vertebrae in Thespesius edmontoni are strongly opistho- 
coelous, and all bear small ribs. An incipient neural spine is present on the 
third, and though the processes increase in size posteriorly, none may be 
said to be especially prominent in size. The centra of the cervical as well 
as those of all the dorsal vertebrae have suffered much from lateral com- 
pression, though they appear to have essentially the same shape and 
proportions as those of T. annectens, and these resemblances are carried out 
in the development of the transverse and neural processes. 
The sacrum is composed of nine coossified sacral vertebrae, with distinct 
spines. All known specimens of the family Hadrosauridae from the Lance 
agree in this respect. 
In the mounted specimen the fifth caudal is shown as bearing the first 
chevron, but from the contour of the lower posterior border of the centrum 
of the fourth vertebra it would seem that perhaps it also may have carried 
a reduced chevron. 
1 Trans. Conn. Acad, of Sc., vol. XI, Jan., 1902, PL 45. 
