PROBLEMS OF HUMAN DENTITION 117 



thousand first molars. And in none of all these, have I met with 

 a vestige of a paramolar cusp. This fact is very impoi'tant, 

 not only as an unexpected peculiarity, but as furnishing a new 

 proof of the correctness of my conception of our first molar as 

 a tooth which was originally a milk molar. This will be made 

 obvious in the discussion of the significance of the morphological 

 phenomena we are about to describe. 



The fact that a paramolar-cusp occurs less freciuently on the 

 third molar than on the second is not difficult to explain. As 

 pointed out above, the opinion advanced by Magitot that a 

 paramolar is always situated in the corner between the second 

 and third molars is incorrect. To prove this, a reproduction was 

 given of a maxilla exhibiting an additional tooth standing in the 

 corner between the first and second molars. This phenomenon 

 induced me to distinguish a Paramolar I and a Paramolar 11. 

 The former, alternating with the first and second molars is less 

 frequent than the latter. By combining all known facts, w^e are 

 able to establish the following coincidence: a rare Paramolar 

 I and a more freciuent paramolar tubercle on the second molar; 

 and further a more fref{uent Paramolar II and a lesser frequency 

 of a paramolar tubercle on the third molar. The relationship 

 between these facts is not difficult to conceive. The first para- 

 molar coalesces with the second molar and the Paramolar II with 

 the third molar. Now the facts enumerated enable us to deduce 

 that the Paramolar I shows a greater tendency to coalesce with 

 its adjacent molar, whereas the Paramolar II shows more incli- 

 nation to remain independent, hence the difference in the occur- 

 rence of the paramolar tul^ercle in the two molars. To get a 

 correct idea of the frequency of the paramolars, it is therefore 

 necessary to compile all the cases of a free Paramolar I with those 

 of a paramolar tubercle at the second molar, and to do the same 

 regarding the Paramolar II and the paramolar cusp on the third 

 molar. The result of this will show that the frecjuency of the 

 two supernumerary elements is nearly the same, with perhaps a 

 small preponderance of the Paramolar II. 



We ha^'e now arrived at the third peculiarity mentioned above, 

 \'iz.: the fact that the paramolar cusp, without exception, is 



