712 



called my attention to an unsymmetrical condition of the sacrum. This 

 seemed of more than passing interest, and led me to examine the 

 vertebral columns of such other specimens of this species as were 

 available. Besides the abnormal animal mentioned, I examined twenty- 

 six others, among which I found a second case of unsymmetrical sa- 

 crum, thus indicating that this variation is not an unusual one. In 

 the twenty -five cases with symmetrical sacra, each sacrum consisted 

 of a single vertebra, on either side of which was a sacral rib carrying 

 the ilium, a condition characteristic for the Amphibia. 



In comparing the vertebral columns, the vertebrae were counted 

 from the anterior end of the column posteriorly, and the positions of 

 the sacrum and of the first caudal vertebra with a complete haemal 

 arch were recorded. I had hoped also to determine the total number 

 of vertebrae in each column, but this proved to be impossible because 

 of the indistinctness of the last few caudal elements. The vertebral 

 columns used in these enumerations were freshly prepared and studied 

 without disarticulating their elements, thus avoiding the possibility of 

 error from displacement or from loss of parts. 



The positions occupied by the sacra and first haemal arches in the 

 twenty-five columns with symmetrical sacra are indicated in the follow- 

 ing table. 



Number of Specimens i ^ ^^ 4 



Position of Sacrum | Vert. 20 Vert. 19 Vert. 19 



Position of First Haemal Arch jVert. 23 Vert. 23 Vert. 22 



It will be noticed that in twenty-one cases out of twenty-five the 

 first haemal arch occurred on the twenty-third vertebra. This coin- 

 cides with the general statement made by Hoffmann ('73 — '78, p. 53) 

 that in this species the first haemal arch occurs on the fourth caudal 

 vertebra, for Hoffmann (p. 51) regards the sacral vertebra as number 

 nineteen, the fourth caudal vertebra thus being the twenty-third in 

 the whole series. In the remaining four vertebral columns of the 

 twenty-five examined, the first haemal arch was on the twenty-second 

 vertebra, a condition that can be designated, to use the terms em- 

 ployed by Bateson ('94, p. Ill), as a backward homocosis, i. e. the 

 condition of an anterior structure assuming the form of one behind it. 



The sacrum in nineteen cases out of twenty-five was formed from 

 the nineteenth vertebra, a condition already recorded for this species 

 by Hoffmann ('73 — '78, p. 51) and by Lucas ('86, p. 562). This 

 clearly represents the typical condition. In the six remaining cases 

 the sacrum was formed from the twentieth vertebra, a variation like- 



