22 IRVING HARDESTY 



in width toward its basal end, the tip of which is its narrowest 

 part; (2) that the actual total width of its apical end in the teased 

 out preparations is about 415 fx or 2.7 times greater than its 

 basal end; (3) that its attached, axial zone varies ver}^ slightly 

 in width, being but very little narrower in the basal end; and 

 finally (4) that, as indicated in figures 1 to 4, its great variations 

 in width occur almost wholly in the width of its thick outspanning 

 zone. As computed from table 1, this zone in the teased out 

 membrane may be about 390 ^ wider in the apical than in the 

 basal end, or the apical 6.8 times the width of the basal end. 

 The same computation applied to the measurements of the 

 sections of the membrane give the outspanning zone in the 1st 

 half turn 3.2 times the width, and in the 3rd half turn 3.7 times 

 the width it has in the 7th half turn. That the differences in 

 the width of this zone do not appear so great in the sections as 

 in the teased out membranes, and that its 1st half turn appears 

 less wide than its 3rd half turn, is explained as due to shrinkage 

 of the apical end, especially in the first half turn, during the 

 preparation of the material for sectioning, and to the differences 

 in the regions of the membrane at which the measurements of 

 its width had to be taken in the sections. The measurement of 

 the 7th half turn especially was in the sections of necessity some 

 distance from the tip of the basal end and therefore could not 

 involve the narrowest part of the outspanning zone. Com- 

 puted from the measurements of the teased out membrane, 

 the width of the outspanning zone in the apical or shortest turn 

 of the coil appears to decrease only about 1 per cent, while in 

 the second whole turn the decrease is 41 per cent, and between 

 the 5th half turn and the basal end, the longest interval between 

 measurements, the decrease is 72 per cent. 



This study of the proportions of the membrane suggests 

 throughout that conclusions as to its actual functional shape 

 based upon its appearance in even the best of sections of em- 

 bedded cochleae are subject to considerable error. 



The thickness of the tectorial membrane had to be measured 

 wholly from the sections. Obviously, the teased out mem- 

 branes could not be used for this. Measurements through 



