624 Journal of Agricultural Research voi. xii, no. i© 



secondary flowers. The remaining loi fruits were borne on flowers 

 bearing a full quota of normal anthers. Of these flowers 54 were primary, 

 40 secondary, and 7 tertiary. While most of the fruits which set on the 

 pistillate plants were of a regular shape, indicating a perfect or nearly 

 perfect set of achenes, those borne on the staminates were for the most 

 part very irregular in shape, as the achenes which set were often few in 

 number and irregularly scattered. Often not more than one or two 

 achenes per flower developed. Where only a few achenes developed, 

 the typical nubbins which are so common in the latter part of the 

 picking season in commercial plantings were produced. These results 

 prove that F. virginiana is a species well on its way toward diecious- 

 ness, and, reasoning from analogy with F. virginiana and F. elatior, 

 it may be concluded that those other American species which produce 

 two types of plants — that is, pistillate and somatic hermaphrodites — are 

 also diecious. 



Recent investigations by Bunyard (6) and Fletcher (74) into the origin 

 of our cultivated strawberries tend to show that they have originated 

 from hybrids of F. virginiana and F. chiloensis, both of which are ap- 

 parently diecious. If this is the case, it raises the question of the origin 

 of our cultivated hermaphroditic forms. A study of the pistil ste- 

 rility in these forms seems to indicate that they may have been derived 

 from males which have varied in regard to pistil fertility. 



Table III shows the relationship between flower position and pistil 

 sterility in 10 hermaphroditic and 4 pistillate varieties. This table was 

 prepared regardless of the degree of setting, whether perfect or whether 

 the resulting fruit was a nubbin, all flowers which set any achenes being 

 put under the heading "Set." The lower horizontal row under each of 

 the two groups indicates the percentage of the flowers of each position 

 which set fruit. It shows the very great increase in sterility from the 

 first flowers to the last in both the pistillate and hermaphroditic forms, 

 being greater in the latter than the former. This is the condition which 

 would be expected if the cultivated hermaphrodites have been derived 

 from males of the wild type, as the males which do set fruit in the wild 

 exhibit a high percentage of their low fertility in the primary flowers. 



The conclusion that the hermaphrodites have been derived from stam- 

 inate forms rather than from pistillate is in keeping with the results 

 found in other species, as Lychnis spp. (55) and the grape {43). A further 

 study of pistil sterility was made in 1 5 other varieties of hermaphrodites 

 and 3 pistillates to determine the relationship between nubbins or irregu- 

 larly set fruit and flower position. 



Nubbins and lack of setting of flowers have been attributed by horti- 

 culturists for the most part to lack of proper pollination or to frost injury. 

 The first of these factors may be eliminated, however, as pollen is usually 

 very plentiful and in a mixed planting, such as the data given in Tables 

 in and IV were taken from, was always abundant and especially so when 



