308 



Belling. 



ovule in a pod. Thus the record for a single pod might read: SO(S)SOO. 

 This means that, counting from the distal end of the pod or ovary, 

 there were: first, a good seed; second, an aborted ovule; tliird, a dead 

 seed; fourth, a good seed: fifth, an aborted ovule; and sixth, another 

 aborted ovule. 



The four parent plants are, to all appearance, constant. No cases 

 of vicinism, segTegation, or mutation have been observed, except an 

 "ever-sportiug"loss of color in the seeds of certain lines of Velvet bean. 



Fig. 3. Pod of Yokohama bean. Natural size. 



This is inherited, and is probably a mutation by loss. I have grown 

 many pedigree lines of Velvet and Lyon beaus, and have found no 

 other variation than the above. I have not seen any insect opening 

 the keel of any Stizolobium in Florida. 



In the cross of the Velvet by Lyon in 1908, th^re was, 1 under- 

 stand, one individual parent Velvet plant; but I do not know whether 

 the pollen came from one or more than one Lyon plant. The parent 

 on one side of the cross of Lyon by Velvet in 1910 was one Lyon 

 plant belonging to a pedigreed line, the ancestor of which was doubtless 

 one of the plants that supjilied poUeu for the cross of 1908. The pollen 

 came from the farm crop of Velvet beans, since no flowers could be 

 removed from the pedigreed lines of Velvet beans, which were under 



