14 MEMOIRS OF THE CARNEGIE MUSEUM 
seventh and succeeding dorsals the antero-posterior diameter of the neural spines 
exceeds the transverse and the extremities of all these vertebree are somewhat ex- 
panded and rugose. All the dorsals of this region exhibit the hyposphene-hypan- 
trum articulation. 
The Sacrum (Plates IV. and V.).— In the present skeleton as in all other fully 
adult Sauropod dinosaurs to whatsoever known genus or species they may pertain, 
there are five vertebree, codssified by their centra and functioning as sacrals by giv- 
ing support to the ilia either by means of so-called sacral ribs or transverse processes 
or by both these elements.. Whether or not all five of these vertebrae should be re- 
garded as true sacrals must remain very largely a matter of individual opinion. 
This matter has already been discussed at some length by the present writer in his 
paper on Diplodocus which formed the first of a series of memoirs, which will con- 
tinue to appear from time to time relating to the dinosaur remains in the collections 
of this museum. 
Whether the number of true sacrals in the Sauropoda be five or less it is evident 
that in those genera of American Sauropods where the complete sacrum is known, 
namely, Diplodocus, Brontosawrus, Morosaurus and Haplocanthosawrus, the number 
of vertebree functioning as sacrals, that is giving support to the ilia, is constant and 
is in no sense different in or diagnostic of the several genera as was supposed by the 
late Professor Marsh. It frequently happens in the case of isolated sacra pertaining 
to young individuals that one or two of the functional sacrals through not having 
been firmly codssified with the three vertebree which, according to the present 
writer's opinion, constitute the true sacrals, have become detached and lost and in 
this manner the number of functional sacrals has been mistakenly reduced to either 
three or four according as the number of detached vertebree was one or two. By an 
unfortunate circumstance this proved to be the case with the sacra of Diplodocus 
and Morosawrus first discovered and described by Marsh. ‘The sacrum of the first 
of these genera was found detached and consisted of three codssified centra, while in 
that of the second (the type of MZ. grandis) there were four codssified centra From 
these circumstances Professor Marsh quite naturally concluded that the number of 
functional sacrals in these genera was respectively three for the former and four for 
the latter and proceeded to make those numbers diagnostic of the genera and fam- 
ilies to which they pertained. Subsequent discovery of more complete material has 
demonstrated beyond a doubt the number of functional sacrals to be five in each of 
these genera as in Brontosawrus. The fragmentary sacra of Apatosawrus and Atlanto- 
saurus figured by Marsh are evidently incapable of furnishing any definite proof as 
to the exact number of sacrals in those genera, which should they finally prove to 
