The Typical Form of the Cochlea and its Variations. 417 



the addition of further whorls to the number that was more remotely 

 inherited. 



This result may seem to contradict that reached above (6), that there is a 

 fair sign of a negative correlation between the absolute length of the basilar 

 membrane or the diameter of the lowest whorl and the index of the second 

 whorl (diameter of the second whorl divided by that of the lowest whorl). 

 But there is really no contradiction. For there is no correlation between 

 the absolute length of the basilar membrane and the index of the basilar 

 membrane. Therefore there may well be an increased rate of curvature for 

 absolutely long basilar membranes (large cochleas) and a decreased rate for 

 relatively long ones. 



12. When the index of the second whorl varies, the index of the diameter 

 of the tube of the cochlea varies in the same direction. This is a practically 

 certain and almost decided correlation. We might well expect a cochlear 

 tube that is thicker than usual to coil less willingly, as it were, and so to 

 give a wide second whorl. If the other dimensions of the cochlea, including 

 the number of whorls, then remained typical, the basilar membrane would be 

 longer for that organ than it usually is for any organ. 



13. The correlation between the index of the basilar membrane and that 

 of the diameter of the cochlear tube at its base is low and uncertain, being 

 only three times the probable error. This would confirm the inferences 

 made in (11) and (12) that in these connections the diameter of the cochlear 

 tube, not the basilar membrane, is the leading variant. 



14. This is further supported by the correlation' indicated between the 

 indexes of the major axis of the oval window and of the cochlear tube. 

 There is no correlation between the former and the index of the basilar 

 membrane. 



We may sum up again by saying that there are two sources of change in 

 the length of the basilar membrane. The chief one is its own absolute 

 increase in length, which appears in a greater number of whorls than usual. 

 The other is the relative increase in the diameter of the tube of the cochlea. 

 There are no other internal variations in the dimensions of the cochlea than 

 these. 



15. In Table IV will be found the ranking of the mammals according to 

 the index of the basilar membrane. The column marked " No." gives the 

 order of succession in which the organs measured are given in Dr. Gray's 

 two volumes, but Dr. Gray himself numbered only the somewhat smaller 

 number of photographs in his books. The absolute length of the basilar 

 membrane is my estimate from Gray's photographs. For the other measure- 

 ments upon which this paper is based, see Gray's two volumes. 



VOL. LXXXIX. — B. 2 M 



