420 DR MATTHEW YOUNG. 



This introduces the question of a reliable base line and Huxley's (5l) well-known 

 dictum on the subject : " No comparison of crania is worth much that is not founded 

 upon the establishment of a relatively fixed base line to which the measurements in 

 all cases must be referred." He chose the basi-cranial axis as his base line because 

 lie thought it showed least variability, but apparently in a uniform series of skulls like 

 the present one under examination the basion-glabella line would be preferable to the 

 basi-nasion line. 



I have been able to obtain the value of the basi-occipito-sphenoid axis in 98 of 

 the same skulls from which the above results were obtained after median section, and, 

 anomalous as it may appear but indubitably proved by the figures, the basi-occipito- 

 sphenoid length, or Huxley's basi-cranial axis, is more variable than the basi- 

 glabella line. 



Perhaps I should explain that by "more variable" I mean that in the series 

 examined it departs more from its mean value than does the basi-glabella line from 

 its mean value. As Pearson (8) explains, we cannot compare the respective vari- 

 abilities of an elephant and a man without making allowances for their difference in 

 size ; or, to state the matter in a different way, it is very evident that a variation of 

 2 inches above or below the mean is very much greater when the mean is equal to 

 10 inches than when it is equal to 40 inches. The magnitude of the coefficient of 

 variation is the true criterion for relative variability, and, according to the evidence 

 supplied by this factor, Huxley's basi-cranial axis is not the least variable length in 

 a homogeneous series of skulls. Of all the chords measurable, the least variable is 

 the greatest length or glabello-occipital length, and I put forward on this account 

 the plea that it should be adopted as a base line if the least variable line is wanted 

 for such a purpose. This length has the disadvantage, however, that its posterior 

 extremity does not come to any definite point on the cranial vault such as the 

 lambda or inion. 



It will be noticed that in his craniotrigonometrical method Klaatsch abandons 

 the glabella-inion base line utilised by Schwalbe and others, and also adopted by 

 himself in some of his earlier research work, and makes use of the glabella-lambda 

 line and plane. He advances various reasons why this line should be universally 

 adopted as a rational base line. There are many arguments against the adoption of 

 the above line as a base line. One of these is the fact that it includes the glabella, 

 which varies considerably in development in different types of crania, but con- 

 siderable variation has also been found to occur in the position of the inion, i.e. 

 the posterior extremity of the glabella-inion line. The chief objection from the 

 point of view of morphological analysis of the cranial vault is that it does not 

 correspond, as the glabella-inion line roughly does, to the sub-cerebral plane. 



According to Bolk (42), the position of the lambda is as variable as that of the 



inion, but my figures from the present series show me that the glabella-lambda is 



liable than the glabella-inion line. Apparently objections may be raised to 



