1909] HILL— POLLINATION IN LIN ARIA 457 



all cleistogamous flowers in respect of the proximity of the organs 

 essential to fertility. It is these cleistogamous flowers I have mainly 

 investigated. My attention was first called to them in 1905, when 

 flowers of this character were found on L. canadensis growing upon 

 the sandstone rocks at Oregon, 111. They were quite inconspicuous. 

 The minute corolla, when pushed off by the enlarging ovary, showed 

 a faint tinge of violet on its upper margin, the main part being color- 



less. 



July 



shoots had disappeared, as well as such flowers as are ordinarily 

 found earlier in the season, if, indeed, they had been formed at all. 

 The plants were generally small, the shortest mostly with simple 

 stems. Some were branched, the tallest about 4 dm high. Since 

 the species is well represented, though not abundant, on the sand 



dunes at the south end of Lake Michigan, there has been an oppor- 

 tunity to observe it each season since, and to note the different stages 

 of development. The plants begin to blossom about the first of May 

 and continue, in some form, the production of flowers till the middle 

 or latter part of July, when the heat becomes too trying for them in 

 the dry sand. On the larger plants there is a gradual diminution 

 in the size of the flowers from the earliest, 6-8 mm loi 

 eter of limb of 8-1 2 mm , till the cleistogamous stage is reached. In 

 some plants of this character, this may occur in the early part of June. 

 It is exceptional to find flowers which are relatively conspicuous dur- 

 ing the later part of the life of the plant, and when found they are apt 

 to be much reduced as compared with the earlier forms. The 

 inflorescence being indefinite, the lengthening of the main stem and 

 branches favors this progressive diminution. Plants that do not 

 exceed 10 to i5 cm usually remain simple and are mainly restricted 

 to cleistogamy. Plants taller than this commonly have flowers 

 adapted to pollination by insects, though it must be rare in the smaller 

 flowers, if done at all, when the limb of the corolla is but 3 or 4 mm " 

 diameter, as Muller found was the case with the small flowers he 



mentions. 



Cleistogamy begins on stems not more than 2 cm high, which may 

 be limited to a single flower at the tip, or perhaps lengthen enough 

 to bear two or three. Flowers will appear on these diminutive stems 

 as early as the larger petaliferous ones on the vigorous plants, the 



in 



