( 367 ) 
the greatest distance from the fronton. With the lower Primates this 
point can as a rule be determined at once, in man, however, and 
especially with juvenile skulls, this is not the case. For it was not 
unfrequently found that a fairly large part of the interior surface of 
the skull in the median plane describes a circular arc with the 
fronton at the centre. Where this was the case the middle of this 
circular arc was always chosen as the posterior point. In what follows 
this point will be referred to as the “Occipiton”. 
The base line extending between fronton and occipiton may at 
the same time be the line of maximum length within the cranial 
cavity and is so in the majority of cases, although not always, 
especially with juvenile skulls. For in man as well as with the 
Anthropoids the frontal wall of skulls where the milk-teeth are still 
present,, shows a more or less distinctly developed concavity. With 
such skulls the line of maximum length consequently lies in a slanting* 
direction through the cranial cavity starting about the middle of the 
frontal bone and ending about the middle of the squama of the 
occipitale. 
Now I shall first point out how in the different skulls I studied the 
Foramen was placed with respect to the base line, after which I shall 
compare its position in juvenile and adult skulls of the same species. 
It was remarked above that this cannot be done by constructing a 
certain angle and determining its value and variations. In general 
it seems to me that in comparative craniology any phenomenon 
roust be studied as little as possible by variations of angular 
values and that the construction of angles has to be restricted. 
For any angle requires three points, the two terminals of the legs 
and the apex. Now a variation in the value of an angle can only 
then be a true criterion of the course of any phenomenon, when 
one i 8 convinced that two of these three points have not changed 
their relative position. And this is hardly to be expected in most 
cases. The position of every point in the skull varies on its own 
account, since on every point a large number of factors have a 
localising influence. Now two points may have a number of these 
factors in common, but besides several others which are different. 
Hence if i n two skulls the angle between three points is found to 
var y, we are not justified to explain the difference by the shifting 
of one of these points only. The inaccuracy inherent in such a method 
I s not entirely avoided, but greatly diminished by constructing pro¬ 
jections on a base line instead of making angular determinations and 
y expressing the course of a phenomenon by the different values 
°f ratio numbers. The degree of accuracy depends with this method 
