2l8 JOURNAL OF CONCHOLOGY, VOL. I4, NO. 7, JULY, 1914. 



Table III. 



bility of more rapid assimilation of food. The figures also show that 

 by far the greater part of the increase in size of the radula is due to 

 an increase in the size of the individual elements, for while the area 

 of the radula increases nearly five times, the number of teeth increases 

 only 0.4 times. It is an interesting speculation as to how far num- 

 ber and size of teeth are related to the efficiency of the organ at 

 different periods of growth, corresponding perhaps to varying dietetic 

 habits. The length of the radula appears to be directly proportional 

 to the diameter of the shell/ but the relations between the other 

 measurements are much more complicated. 



We may next consider the form of the individual teeth. Here 

 interest centres chiefly about the third lateral and the first marginal 

 denticles, i.e., the third and fourth teeth from the central tooth. 

 Throughout the series the first two teeth are definite laterals, i.e., 

 they have three cutting points. Similarly the fifth and succeeding 

 teeth to the margin of the radula are truly marginal in type, i.e., they 

 have only one cutting point and no reflexion. The third tooth (third 

 lateral) on the other hand, may have one, two, or three points, and 

 the fourth tooth (first marginal) either one or two. The figure shows 

 typical examples of the first five teeth in radulse from different sized 

 snails, all drawn to the same scale with camera lucida.'-^ The third 

 lateral in the smallest specimens has a mesocone only (type E), and 

 is typically marginal in form ; endocone ^ and ectocone successively 

 develop until a stage is reached in the largest specimens (type A), 

 in which the tooth is of a definite lateral type. The first marginal 

 correspondingly develops an endocone in some of the largest speci- 

 mens. It will be remembered that a first marginal tooth with a 

 tendency to assume lateral characters has been noted as characteristic 



1 The length of the radula may be calculated pretty accurately by dividing the major dia- 

 meter of the shell by 2.6. 



2 H. V. Ihering (in C. A. Westerlund, Fundamcnta Malacologica, 1892, p. 56) says : " I am 

 wholly averse to the making of drawings of teeth with the camera lucida It is unnecess- 

 arily precise in details " ! 



3 I use a conventional and intelligible nomenclature. More properly, perhaps, the lateral 

 teeth in Hyalinia have no endocone, but a bifid mesocone. 



