36 THE MEASUREMENT OF VARIATION. 



limbata. Four hundred individuals were examined, 

 and it was found that, as regards the distribution of the 

 total number of teeth, the left total fell into a curve of 

 Pearson's Type I, whilst the right total was of Type IY. 

 However, by dropping out a single individual from the 

 series, it was found that the curve was thrown from 

 Type IV to Type I. As Miss Hefferan points out, this 

 raises a serious question as to the biological importance 

 of the distinction between Pearson's Type I and Type 

 IV.* 



* Should any further information regarding these asymmetrical 

 curves be desired, the reader should consult Professor Pearson's 

 memoir on the subject, or, if he is not a mathematician, then a 

 recently published book by Davenport on " Statistical Methods,"* 

 and also a paper by Duncker on ' ' Die Methode der Variations- 

 statistik " f may be referred to. Both of these are said to be written 

 specially for biologists. I must mention my special indebtedness to 

 Duncker's paper, which has been drawn upon freely in writing the 

 last few pages of the present chapter. 



* New York and London, 1899. 



fArch. f. Entwickelungsmechanik, Bd, viii. p, 112. 



