MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



177 



One peculiar effect of the occipital flattening is observed in the horizontal circumference. In 

 certain of these skulls (H. 4, H. 10, and H. 47) a curious difficulty has been encountered concerning 

 the horizontal circumference. It is prescribed that this circumference, which is supposd to be the 

 maximum, be taken on a line passing above the supraciliary ridges and through the maximum 

 occipital point ; thus the posterior segment of the circumference encircles, so to speak, the posterior 

 end of the maximum length. But in these skulls the line indicating the greatest circumference 

 passes high up toward the obelion, and is drawn through so high a plane of the skull that the 

 greater breadth of the skull at points below that plane more than compensates for its slightly less 

 length ; therefore the maximum circumference does not lie in the same plane as the maximum 

 length. 



Fio. 29. Occipital depression, right lateral. 



Again; suppose that we take a skull of any ordinary shape and paint a line around it in the 

 horizontal plane of its greatest length. If we then look downward upon the vertex of the skull 

 we shall hardly see the line at all, because it corresponds so nearly to the outline of the skull in 

 normai verticals; but if we take one of the deformed skulls in question and paint a line correspond- 

 ingly related to the maximum length and then look down upon the skull, we shall see painted 

 upon it an ovoid figure which coincides with the outline of the skull only at its posterior extremity. 

 This is owing to the fact that the most protuberant regions of the cranial parietes are situated 

 much below the horizontal plane of the greatest length. 



In these cases both the maximum circumference and the circumference around the maximum 

 occipital point have been recorded, although it has been a matter of great difficulty to determine 

 exactly the maximum circumference, and a series of measurements of the same made at long inter- 

 vals of time would probably show considerable variation. 

 S. Mis. 109 12 



