6 16 Journal of Agricultural Research voi. xii. No. lo 



In F. americana, lo, 15, and 20-stamen arrangements are common. 

 The 20-stamen arrangement is the same as that described for F. vir- 

 giniana and the cultivated varieties. When, however, a decrease 

 below 20 to 15 takes place, it is due to the loss of the short filamented 

 middle whorl of antipetalous stamens (fig. 2, /) — further proof that 

 this is truly an inner and not a middle series. The next series to be 

 lost is that consisting of the inner long filamented antisepalous stamens, 

 thus leaving the parapetalous stamen arrangement (fig. 2,e). This seems 

 peculiar, in view of the fact that they are older than the parapetalous 

 stamens and therefore should remain longest. This might be consid- 

 ered as proof that — 

 the parapetalous stamens must be regarded as abnormal supernumary parts, 



as Rydberg {34, p. 11) considers them. The genus Fragaria differs 

 from other species of the Potentilleae in this respect, as the more usual 

 order of loss is first, the parapetalous, followed by the antipetalous 

 stamens {34, p. 11), the long filamented antisepalous stamens being 

 the most permanent. When stamens are dropped in Fragaria spp., 

 they are lost completely and do not form the staminoids or partially 

 developed stamens which are found in the pistillate flowers, so that a 

 decrease in stamen number can not be considered as a step toward 

 dieciousness. 



Staminodia. — Typically F. virginiana and, as will be shown later, 

 some other species of strawberry are diecious, although the separation 

 into strictly staminate and pistillate forms is not complete. The flowers 

 of pistillate plants bear staminodia showing varying degrees of develop- 

 ment (fig. 3, 11-16), which never, as far as I have observed, produce 

 pollen.^ The staminate plants develop pistils which, as far as outward 

 appearances are concerned, are normal, but which do not set fruit. As 

 a result of this inconiplete separation of the sex-bearing organs, there 

 are variations in the stamen and pistil condition on individual clones 

 and also on the flowers of an infloresence within a clone. 



The most common condition in the staminate plants is that in which 

 all of the flowers produce normal stamens bearing good pollen. Occa- 

 sionally clones are found in which the first flower bears only staminodia 

 in place of the normal stamens. In flowers of this type the pistils are 

 practically always fertile and produce normal fruits. On the other 

 hand, the primary flower may produce normal stamens and no fertile 

 pistils while one or both of the secondary flowers may be of the pistillate 



1 A possible case of pollen production in a pistillate variety is that of the Crescent. Plants of it which 

 I have observed 'are strictly pistillate, although producing very large prominent staminoids (fig. 3, 16)- 

 which are entirely devoid of reproductive tissue (PI. 36, A). Castle (7 ,p. 150) states that in England "it 

 produces perfect flowers and sets its fruit most readily, cropping heavily in favorable seasons." As most 

 other English varieties are hermaphrodites, it is very possible that large crops might set as a result of cross- 

 pollination, and that the presence of the extremely large staminoids has been misleading with regard to 

 the exact sexual condition of these plants. 



Fletcher (75, p. 132) also states that Crescent may vary in stamen condition becoming "a true stami, 

 nate on rich soils," but gives no ftuther evidence on the point. 



