112 ESCHRICHT AND REINHARDT 



means the upper part of the process looks as if it were shghtly expanded, and the whole margin 

 becomes sigmoid. 



As the arches of the dorsal vertebræ leave their horizontal position, they lose also their 

 imbricate or overlapping position relative to each other, and every trace of the posterior articular 

 processes has disappeared in the ninth dorsal vertebra. 



The anterior articular processes are developed in inverse proportion. We find their first 

 traces in the second or third dorsal vertebra in the form of a small knob quite externally on the 

 anterior margin of the transverse process, and they are placed here close outside the posterior 

 articular process of the preceding vertebra. In the fourth and fifth dorsal vertebræ they are 

 already rather prominent on the anterior margin of the transverse processes, causing the exterior 

 half of this margin to become very concave ; in the sixth and seventh they appear in the form 

 of a pair of wings quite inwardly on the anterior margin of these processes, causing this margin 

 to become imiformly concave throughout its entire length ; in the eighth and ninth they are 

 removed upwards to the most external part of the neural arch, and at the same time they are bent 

 in an outward direction, so that the exterior lateral surfaces of the vertebræ are made to form a 

 deep excavation in the longitudinal axis of the vertebral body. But that part of the articular 

 processes which is bent in the manner described, assumes in these verteTjræ the character of an 

 extremely rough osseous protuberance (perhaps the so-called processus mamillaris). In the 

 tenth and eleventh dorsal vertebræ these articular processes have ascended upwards on the 

 branches of the neural arch, and have accordingly been separated from direct connection with the 

 transverse processes. The rough osseous protuberance has become more narrow, lengthened, 

 and points more upwards than outwards. They now appear like wing-like, vertically placed, 

 osseous plates, and the extended and rough protuberance is changed into the superior rough 

 margin of these plates. At the same time they clasp the spinous process of the preceding 

 vertebra, though never so closely as in the rorquals. 



Of the difi'erent processes of the dorsal vertebræ, the transverse processes unquestion- 

 ably deserve the greatest attention. These processes differ much from one another. In the 

 five foremost dorsal vertebræ they appear as long and narrow prolongations, pointing very much 

 forwards ; in the five middle ones they are short and flat, and form right angles with the 

 bodies. They become continually broader, as they succeed one another from before backwards, 

 though, especially at their inner and outer extremities, recalling the form of an hour-glass 

 (having its inner and posterior angle worn off) ; finally, in the three hindmost, they resemble 

 a pair of long, flat bone-plates with direction exactly transverse and horizontal. We have 

 already mentioned, that they have their origin in the same points as the lateral branches of the 

 arches in the foremost dorsal vertebræ, whereas in the very hindmost ones they arise only very 

 close to these arches, but independently from the upper part of the lateral surfaces of the 

 vertebral bodies. 



As to the differences between the transverse processes in the five foremost dorsal vertebræ as 

 compared with one another, we may refer our readers partly to the three woodcuts, pp. 107, 108, 

 and 109, as far as the first dorsal vertebra is shown there, together with the cervical vertebræ, 

 and partly to the illustration given at Plate II, fig. 3, copied from a photographic representation 

 of the cervical and the anterior dorsal regions of our twenty-two and one third feet long skeleton. 

 In this figure the transverse processes, as far as they are visible, are in the right side marked 

 with '3, '3, '4, and in the left side with Ij, 2', W, b'_. They were not yet ossified to their outer 



