^"'igi^^'l SnuFiii.UT, Osteology of Harris's Cormoranl. g3 



vertebrae, although I do not think I did, 1 determined to test it 

 with some rigid long hone ol' the skeleton, and selected for the 

 purpose the hiuncrtis. Dr. (iadow found tliis bone in his skeleton 

 of Nannoptcritm havrisi to liavc a Irngth of loi niillinielers (/(h'. 

 cit.. Tab. A, p. 171). 



The humerus in the skeleton of No. 19,720 of the U.S. National 

 Museum collection, a large specimen of Harris's Cormorant, 

 measures in total length fully iii millimeters ; in the one numbered 

 19,719, III millimeters ; and the same bone in No. 19,721, only 

 98.5 millimeters ; so that the lengths of the three himieri would 

 give us an average of io().8 mm., or over 5 millimeters longer than 

 the big male harrisi that Dr. Gadow had before him. 



The humerus of the specimen of P. carbo Dr. Gadow had had a 

 length of 165 millimeters, and that bone of an individual of the 

 same species in the Coll. U.S. Nat. Mus. (No. 18,851) has identically 

 the same length : so I am inclined to think that the long bones 

 in the pectoral limbs of these birds exhibit, when adult, but a 

 very slight amount of variation. Further on I shall have some- 

 thing more to say about these measurements ; the question was 

 only introduced here in order to demonstrate such value as they 

 possess in the matter of the total length of the trunk skeleton in 

 Nannopterum harrisi. 



Harris's Cormorant stands among the largest of the Phula- 

 crocoracidce in the world's avifauna, and the fact of its having 

 lost its power of flight has, pari passu with the atrophy of its 

 pectoral limbs and certain parts of its sternum and shoulder- 

 girdle, resulted in an augmentation in size and strength of most 

 of the bones in the remaining parts of its skeleton. In this 

 category the vertebral column and pelvis hold a prominent place, 

 while the skeleton of a pelvic limb is, actually as well as relatively, 

 better developed than that of any other Cormorant at present 

 known to me. 



In the cervical section of the spinal column the first 18 vertebrcB 

 are without free ribs. The igth vertebra is free, and supports a 

 pair of long, slender ribs. These have elongate, well-developed 

 epipleural appendages upon them ; but the ribs themselves are not 

 articulated with the sternum through costal ribs. In this 

 Cormorant the 20th and 21st vertebrce fuse solidly together to form 

 one piece, the ribs on the 20th being longer and stouter than the 

 pair on the 19th, though they still fail to connect with the 

 sternum by means of costal ribs. It is interesting to note that 

 in both pairs of these ribs their lower free extremities are knobbed, 

 just as though the usual provision had been made for articulation 

 with costal ribs ; but the latter, as I say, are non est, neither are 

 the vestiges of facets to be seen on the costal borders of the sternum 

 for them, indicating that these two pairs of ribs, in the ancestors 

 of this Cormorant, did so articulate with the sternum through the 

 intervention of such costal ribs. 



As stated above, the 215^ vertebra is completely fused with the 

 20th, even the large quadrate neural spines being included in the 



7 



