412 



Mr. H. J. Watt. 



last have also a high positive correlation with one another {vide Table I, 

 Note 1). The correlation with the length of the head and body is of the 

 same order (vide Note 2). 



Table I. — The Cochlea of Mammals. Correlation Coefficients (from 

 K. Pearson's formula). 51 or 52 organs (42 for the length of the 

 basilar membrane). 





Diameter of the 

 lowest whorl. 



Length of the 

 basilar membrane. 



The number 

 of whorls. 



Diameter of second whorl 



Length of the basilar membrane 



Major axis of the oval window 



Diameter of the tube of cochlea 



Diameter of second whorl 



— - per cent. . . . 



Diameter of first whorl 



•970 ± -OOG 

 1 -963* ± -008 1 

 \ -954 ± -009 J 

 •914 ± -015 

 •806 ± -033 



- -417 ± -077 



■954 ± -009 

 •931 ± -014 



■812 ± -035 

 •760 ± -044 



- -349 ± -091 



•088 ± -092 

 •070 ± -092 



•155 ± -101 



•035 ± -094 

 *195 ± -090 



•000 ± -094 



* For the length of basilar membrane in the first two whorls. — 



(1) r for diameter of tube of cochlea to major axis O.W. = - 828± '030; 



(2) r for diameter of lowest whorl to length of head and body = '864 ± '024. 



2. The series of correlations for the length of the basilar membrane 

 appears in the second column of the Table. It will be seen that it is 

 parallel in its degrees to the series appearing in the first column, but that 

 the values of the second column, with the exception of the reciprocal pair, 

 are all smaller than those of the first, while the values of the probable 

 error are, of course, larger. If the length of the basilar membrane and the 

 diameter of the lowest whorl were perfectly correlated with one another, 

 this difference between the two columns would mean that my measurement 

 of the length of the basilar membrane is subject to a variable error that is 

 probably greater than the error made by Dr. Gray in his direct measure- 

 ments, but that is still not disturbing. But, as we shall see, the relative 

 length of the basilar membrane is subject to special variations in relation to 

 the number of whorls of the cochlea (vide below, B). These fluctuations 

 would account for some part of the difference between the two columns. In 

 any case, my measurements of the length of the basilar membrane may be 

 taken as good. 



3. There is no correlation at all between the number of whorls and the 

 diameter of the lowest, or even of the second, whorl, or the major axis of the 

 oval window, or the percentage relation of the second whorl to the first. The 

 other two correlations, with the diameter of the tube of the cochlea and with 

 the total length of the basilar membrane, are too small to be of any 



