GRAMINEAE 531 



related species (i.e. those sometimes included in the" genus Aegilops) agree with 

 T. ovatum in their manner of flowering, but Kirchner only rarely observed flowers 

 still open after the proper period of anthesis. 



3063. T. cylindricum Ces. Pass, et Gibb (= Aegilops cylindrica Host). 

 Hildebrand states that the homogamous flowers of this species, which is indigenous to 

 Hungary, Slavonia, &c., open towards midday. After a slight divergence of the 

 glumes, the dehiscent anthers tip over downwards; simultaneously the stigmas 

 project laterally a little at the base of the flower, so that self-pollination can take 

 place. The stigmas remain receptive and the flowers remain open for some time 

 after the anthers have fallen, so that cross-pollination is now possible. (Loew, loc. cit.) 



1014. Agropyron J. Gaertn. 



3064. A. repens Beauv. 



Visitors. MacLeod observed a Muscid (Spilogaster sp.). 



3065. A. caninum Beauv. The flowers of this species are protogynous and 

 chasmogamous ; the glumes diverge and the stigmas project at 8 a.m., the anthers do 

 not dehisce until towards midday, and scatter their pollen when they protrude from 

 the glumes. The pollen-grains are pyramidal in form, yellowish-white in colour, 

 almost smooth, on an average 50 fi in diameter (Warnstorf, Verb. bot. Ver., Berlin, 

 xxxviii, 1895). 



1015. Secale L. 



3066. S. cereale L. (Sprengel, * Entd. Geh.,' pp. 79-80.) Hildebrand 

 states that in this species the anthers first project between the still partially closed 

 glumes ; when they have protruded to their bases they tip over, a part of the pollen 

 falling out at the same time, so that the stigma of the same flower cannot be dusted. 

 Later on the glumes diverge widely for some hours, and the stigmas now bend 

 forward for the first time. Meanwhile the anthers have dehisced more fully and the 

 pollen is gradually shaken out, reaching the stigmas of other flowers more easily, 

 because the anthers are situated below the stigmas. Cross-pollination is therefore 

 favoured, but autogamy not excluded {cf. Loew, ' Bliitenbiol. Floristik,' p. 372). 



Godron and Kerner state that the flowers open between 6 and 7 a.m. The 

 former says that the anthers tip over immediately under favourable conditions, so 

 that self-pollination cannot take place. In colder regions (northern latitudes or 

 high mountain levels), cleistogamy or, according to Hansgirg, pseudo-cleistogamy, 

 occurs. 



Askenasy (Verb. nat.-hist. Ver., Heidelberg, New Sen, ii, 1879) states that after 

 the divergence of the glumes the stigmas first bend rapidly downwards, and the 

 stamens elongate, while the anthers dehisce simultaneously. Autogamy is, however, 

 not completely excluded, but xenogamy or geitonogamy by means of the wind is 

 generally effected. 



Rimpau says that single flowers open at a temperature of 1 2^ C. The period 

 of anthesis is not so limited as Godron asserts, but is distributed over the whole 

 morning. Cross-pollination is ensured by projection of the stigmas for some time 

 after the closing of the glumes. The self-sterility of the species described by 



M m 2 



