37 



are entirely confirmatory, excepting concerning the theory of irregular 

 segmentation which shall be discussed later. They call for a great 

 change of thought on the part of many who have accepted Rosen- 

 berg's view, and even raised them to a criterion. There is no moving 

 of the ilium towards he head in repetition of phylogenesis, no releasing 

 of a vertebra to attach itself to one before it, no delay of its motion 

 from ancestral tendencies, no acceleration of it in anticipation of the 

 future. The variation which occurs may well be called variation round 

 a mean. Bardeen has shown that an increase and a decrease of the 

 prsesacrals is equally likely. I had looked upon concomitant variations 

 in ditierent regions as frequent. Adolphi shows that they are the 

 rule. His observations as well as those of Ancel and Sencert accord 

 with the views held by Tenchini, myself and others that the error in 

 development having once occurred the organism tends to correct it 

 as much as possible. 



This is the place to explain my method of measurement which 

 my French colleagues criticize. They measure the whole spine. I 

 measure only to the sacrum. If the parts are normal and fresh so 

 that the intervertebral discs are not shrunken I measure each region 

 including the discs. If they are not fit to use I then measure the 

 anterior height of the vertebra? beginning with the 3rd and of course 

 compare each set of observations with averages of similar ones. It is 

 elementary that parts with an important physiological function are far 

 more stable than those with a very subordinate one or with none at 

 all. The functions of the praesacral spine are very important; they 

 are to support the body and to form the thorax. Below the pro- 

 montory the condition is essentially different. The first three sacral 

 vertebrae are indeed important in forming the upper part of the true 

 pelvis, while the lower sacral and all the coccygeals are of compara- 

 tively little value. Not only is their number uncertain but also their 

 size and shape. This is why I measured my spines as if the prse- 

 sacral region were to be considered by itself. This may not be correct 

 from a morphological stand -point, but it is so from a physiological 

 one; and we must recognize that modifications in size and shape by 

 which an error of development is to a certain extent corrected have a 

 teleological bearing. In the spines presenting a prsesacral too many 

 the changes of this kind are much less evident than when there is a 

 prsesacral too few, because in the former case a sufficient thorax is 

 assured. If there are but eleven thoracic vertebrae they very often 

 are longer than usual. Often there is a cervical rib. If the last rib 

 be rudimentary the penultimate is unusually long; if there be an irre- 



