1894.] ormTATiiEE of the sptntb in fishes. 99 



the second of the two furnished by Prof. Sutton. The entire 

 column consists of 48 vertebrae, and its most noteworthy feature 

 is a single flexion involving the 30th to the 35th of the series. 

 These are so modified (fig. 4 a) as to form an arch, of which the 

 33rd (c. 33) is the keystone. The ventral compression of the 

 33rd vertebra of this specimen is more marked than that of any 

 similarly modified vertebra with which I am familiar; and in 

 accordance with this and the corresponding adaptive shelving of 

 the anterior faces of the 34th and 35th, the succeeding vertebrae 

 must in life have been disposed at a sharp angle to those in front 

 of them. 



It is characteristic of the specimens which I have thus far de- 

 scribed that where sinuosity occurs synostosis is uneffected, but 

 inasmuch as in the example now under consideration vertebrae 

 nos. 31 and 32 (cf. fig.) are partially united, that so far bridges 

 over the gap between the sinuous and compressed types. This 

 union is seen to be the outcome of an extension of the right 

 base of the 31st haemal arch (a.h. 31), that structure, as it 

 were, having welded together the two vertebrae. In cor- 

 relation with this there have arisen a series of displacements 

 involving only the right side, rendering it at first sight apparent 

 that the 35th and 36th haemal arches are double. This is in reality 

 not so, for detailed analysis shows that the right half of the S2nd. 

 haemal arch had become shifted back and confluent \vith the body 

 of the 33rd vertebra, while the coi-responding halves of the 33rd and 

 34th arches had become similarly shifted and co-ossified with the 

 vertebrae (34th and 35th) next in order of succession behind. 

 The two halves of the 36th haemal had, in sympathy, but insignifi- 

 cantly united beneath the haemal canal, and the right half of the 

 35th had entirely disappeared. 



The arches of the remaining vertebrae of this specimen are normal; 

 but those of the distorted region present, in addition to the features 

 already described, an irregular lateral disposition, those of the 31st 

 and 32nd especially being so modified as to conform in end view 

 to the limbs of an S-shaped curve. 



There can be httle doubt that the synostoses, compressions, and 

 sinuations afore described are, as HjTtl surmised for the first- 

 named, congenital in origin. As remarked at the outset {ante, 

 p. 95), it is generally the custom to regard the causes producing 

 congenital cin-vature of the spine as unknown. This may be so for 

 lateral curvature, but concerning the vertical variety herein dealt 

 with a consideration arises. The facts which I have recorded 

 appear to me to point towards the conclusion that divergent as the 

 conditions of sinuation and compression with or without co-ossifica- 

 tion appear, they are in reality the opposite effects of one and 

 the same disturbing influence; and, indeed, the indication of a 

 sinuous arrangement in the compressed type (fig. 6) suggests that 

 they are perhaps even more closely related. In both there results 

 an approximation of the opposite spinal extremities, and, in relation 

 to the vertebrae of each individually disturbed series, of the opposite 



7* 



