M. li. Crane 5 



The plants raised both years were from seeds from the same fruits 

 gi-own in 1912, and it is conceivable that the seeds which produce the 

 compound types may lose their vitality earlier than the simple type. 

 It is interesting however to note that out of 20 F:, families laiscd only 

 two small families (2!) plants) were heterozygous for compouiid inHores- 

 cence, and 24 were simple and 5 compound. 



The compounil inflorescence tyjjes are very distinct from the simple, 

 but there is considerable variation in them, some being much larger 

 and develop much more foliage than others, but the primary dichotouious 

 branching system is common in all. 



Fruit. 



The shape of the fruit of 102 plants was recorded. They may be 

 divided into three primary ftjrms — conical, round and long. All these 

 however can be further subdivided as : conical and pointed conical 

 (heart-shape), spherical round and compressed round, full longs and 

 pyriform long (longs with a constriction at the proximal end), etc. 



The analysis of shape is a difficult problem, e.specially with allied 



forms, and undoubtedly many factors are concerned in the shape of the 



fruit. The family was not sufficiently large to justify the classification 



of all the various forms. If however length alone is taken, definite 



results appear to be obtained from this family; by classing the various 



forms of conical and round as short, and the full and pyrifoini longs as 



loiKj, We obtain appro.ximately a .3:1 ratio, the longs being recessive'. 



The numbers obtained are: 



Short Long 



78 (conical -f round) 24 



Expectation 76:5 25-5 



Although short and long fruits and simple and compound inflores- 

 cence both behave as unit characters, the long fruit and compound type 

 of inflorescence being the recessives, the proportions recorded were not 

 normal, as there is a considerable divergence from the usual 9:3:3:1 

 ratio. The two small F^^ families, which were heterozygous for the 

 inflorescence, were also heterozygous for fruit shape. If we combine 

 them with the F« the results are as follows (^4 representing the factor 

 for the simple typo of inflorescence, and B the fiictor for short fruit) : 



AB Ab aB ab 



Actual numbers 91 : IG ill : 13 



Expectation of D : 3 : :5 : 1 ratio 7362 : 2i 5i : lii.'H : SiS 



' There appeared in two /•', funiilii's a distinct flattened round, which is apparently 

 recessive to long. 



