OOTOBEK 14, 1910] 



SCIENCE 



525 



■0000 -471* -8165 ■ -9428 



DlAQEAM 3. 



as 4, 6 and S are conveniently designated as 

 " even " while those with the numbers 5, 7, 9 

 and so on are tersely described as " odd." 



The suggestion that these two types of 

 locules may differ in their capacity of develop- 

 ing to maturity is not the product of the 

 mathematician's fancy, but the working 

 hypothesis sequent upon a large mass of bio- 

 logical observations. The biological reasons 

 for determining whether there is a selective 

 elimination of fruits with a preponderance of 

 " odd " locules are two. 



First : I knew from large and numerous 

 series of statistical data that there are in some 

 species of plants differences in the relative 



0000 -4714 -3165 

 Diagram 4. 



■94284- 



numbers of " odd " and " even " locules in the 

 matured fruits. 



Second: The ovules in the flowering plants 

 are generally borne on the two margins of a 

 carpellary plate. The ovary is formed by the 

 fusing of the two margins of the embryonic 

 plates, either with each other or with the 

 margin of the adjoining plate. So far as I 

 am aware the development of the ovary of 

 Staphylea has not been investigated, but judg- 

 ing by analogy witli other forms, it would 

 seem that each locule represents a single car- 

 pel with two ovule-bearing margins fused 

 together. 



K this be true it is clear that a carpel 



