326 J. ARTHUR HARRIS 



quite dissimilar horticultural varieties and comprising 6,233 and 

 8,018 pods maturing 18,575 and 21,231 seeds respectively, leads 

 to the following conclusions for Phaseolus: 



1. No relationship between the position which a seed occu- 

 pies in the pod and its liability to parasitization has as yet 

 been demonstrated. 



2. The relative number of seeds which are parasitized J in- 

 creases from pods with the lower to pods with the higher num- 

 bers of ovules. That is to say, the percentage frequency of 

 seeds which are injured by parasites is higher in pods with the 

 larger numbers of ovules. Of course this percentage is based 

 upon the total number of seeds matured, not upon the total 

 number of ovules 4 originally laid down. 



3. The same relationship holds for the number of seeds ma- 

 tured per pod and the incidence of the parasites. The seeds 

 are relatively more extensively infested in pods maturing larger 

 numbers of seeds. 



4. It is difficult to differentiate and to measure the inde- 

 pendent influence 6 of these two characters of the ovary upon 

 the incidence of the insects. Apparently the number of seeds 

 matured has some influence independent of that of the number 

 of ovules per pod with which it is correlated. Since this rela- 

 tionship seems to be slight, it is probable that the number of 

 ovules laid down is the primary factor. 



The simplest hypothesis in explanation of the observed rela- 

 tionships would seem to be that in the young pods size is cor- 

 related with the number of ovules formed and the number of 

 seeds which are beginning to develop, just as they are known 

 to be in matured pods of other forms, 6 and that in consequence 

 the maintenance of a foothold and oviposition are easier in the 

 larger ovaries. This is, however, merely a suggested hypothesis 

 which must be confirmed or disproved by actual behavior studies. 



8 In any future studies of this kind the seeds should be so preserved that the 

 number of insects, if more than one, which emerge from each seed may be deter- 

 mined. 



4 The number of ovules per pod includes the number of seeds actually matured 

 and the number of abortive seeds. 



5 Since number of seeds matured and number of ovules formed per pod are posi- 

 tively correlated (Harris, J. Arthur, Arch. f. Entwicklungsmech. d. Organism, 35, 

 500-522, 1912) a relationship between either of these characters and the incidence 

 of weevils would necessarily result in some relationship between the other char- 

 acter and frequency of parasitization, even though there were no direct causal 

 relationship between them. 



6 Harris, J. Arthur, Bot. Gaz., 50, 117-127, 1910; loc. cit., 53, 204-218, 396-414, 

 1912. 



