hatcher: diplodocus (marsh) 37 



of terminating below the hsemal canal in a single compressed spine, as in the chev- 

 rons of the anterior cauclals, these spines are expanded and present anterior and 

 posterior branches. The posterior branches are at first the longer of the two, but 

 they soon become snbequal in length. According to Osborn all the chevrons pos- 

 terior to the thirteenth caudal are open above the hsemal canal. The anterior and 

 posterior branches of the anterior branched chevrons surround a long median open- 

 ing which is confluent with the haemal canal. While in this region the inferior 

 branches of the chevrons are coalesced at their extremities, posteriorly they are re- 

 duced to long slender rods closely applied to, but entirely separate from one another 

 throughout their entire length. Each chevron is united more closely to the posterior 

 of the two vertebrae with which it comes in contact than with the anterior. Ac- 

 cording to Osborn the sixteenth chevron is firmly coalesced with the centra of the 

 eighteenth caudal in the American Museum skeleton, and among the caudals found 

 with No. 94 in our collections there is one also bearing a codssified chevron which 

 compares well in size and form with the eighteenth in Osborn's series. This would 

 seem to indicate that this is a constant feature. Moreover this is just that region of 

 the caudal series which would come in contact with the ground when in life this 

 animal assumed a tripodal position. It is to this end no doubt that the chevrons 

 of this region have been so modified and their union with the caudals made more 

 complete in order the better to resist the impact brought to bear at this point by the 

 superimposed weight of the tail and body, more especially when the animal assumed 

 a tripodal position. 



The Vertebral Formula. — From the above description it will be seen that the ver- 

 tebral column of Diplodocus contains about seventy vertebrae. The following for- 

 mula indicates the number of these belonging to the different regions as indicated 

 by the material at present available for study, viz., cervicals 15, dorsals 11, sacrals 

 4, caudals 35 to 40 or even more. While these figures cannot be taken as abso- 

 lutely correct, they cannot be far wrong. Whatever change either in the absolute 

 number of vertebrae in any single region, or of the vertebral column as a whole, 

 may be necessitated by the future discovery of more perfect material, we may be 

 perfectly sure that the relative proportions of the several regions as now understood 

 will not be materially changed. There can be no doubt that the -pre- and postsacral re- 

 gions were subequal ; that the centrum was the center of power and nodal point in the ver- 

 ebral column; that there ivere no true lumbar s; that the dorsals luere for the most part 

 jhort and few in number, resulting in an abbreviated dorsal region; that the cervicals 

 loere elongated and more numerous than the dorsals, resulting in an elongated cervical 

 region. The elongated and increased number of cervicals, shortened and reduced 



.-/'' 



