127 



July in Avliicli the calyx remained closed over the dwai'fed corolla. The 

 only change in striictni-e noted was that the five shorter stamens bore 

 aborted anthers, and that the pollen-bearing anthers were in contact with 

 the stigma. Contrary to what some writers state the stamens of 0. 

 Stricta are often dinorphic. Tlie self-pollenation of the normal flower is 

 accomplished by the corolla closing after exposure, and pressing the 

 anthers against the stigma. 



One of the most interesting changes in structure from a conspicuous 

 to a cleistogamous flower is seen in the violet. The shoAvy flowers are so 

 constructed that the honey-bee is the only insect that I know to be of 

 service in its fertilization, and only a part of. the anthers are called into 

 use by it. To reach the spur in which the nectar is stored, the bee, after 

 it settles, has to reverse its position, and force its tongue between the two 

 appendages on the lower stamens. In doing this it comes in contact 

 with the stigma and at the same time is dusted with pollen from the 

 appendagrd stamens. The anthers of the other three stamens do not aid 

 in supplying the bee with pollen, and seem to be of very little if any use 

 to the plant. In the concealed flowers, they are aijorted. The pistil, of 

 the cleistogamous flowers of ^'iola Sti-hitd Ait., is declined, so as to bring 

 the stigma against the end of the ovary, and in contact with the two con- 

 niven.t anthers. Two appendages grow from the fertile stamens, just 

 below the anthers, that are expanded so as to cover the anthers and the 

 whole of the pistil. 



r. striatii continues to produce showy flowers longer than many other 

 species, and as a consequence its concealed flowers come in summer. 



Yiohi piihfsrciis Ait. develops a few yellow flowers in early spring. It 

 continues to grow until August, and as it grows, concealed flowers are 

 developed in the axils of the leaves. 



The aliruptufss of the change from a showy to a cleistogamous flower 

 was beautifully shown on a plant of Iiiipatiens hiffora that produced a well- 

 developed, cons]ucuous flower on one branch of a peduncle and a concealed 

 flower en the other branch. 



The fact that the stigma of Tecoma radicans returns to its former posi- 

 tion in two hotirs after it has been changed in response to an irritant, 

 imless the irritation has been accomplished by pollen of a certain quality, 

 shows that the process of fertilization begins within two hoiu's after the 

 right kind of pollen has been applied, and that the stigma is endowed 

 with remarkable selective power. The whole process suggests the shad- 



