AUTOGAMY BY MOVEMENTS OF STAMENS. 889 



flowers should visit tliem in quest of honey. lu a short time the stamens opposite 

 the petals raise their anthers to the same height as the stigniatic tissue of the 

 divergent styles; but the filiform filaments slope away from the axis, so that there 

 is always some interval, however small it may b&, between anthers and stigmas, and 

 there is still no autogamy. It is not till the last moment, when the flowers begin 

 to close, that the stamens opposite the petals incline towards the centre of the 

 flower, and, laying their anthers upon the stigmatic tissue, cover it with a quantity 

 of their still abundant store of pollen. In most of the Alsinere, of which we are 

 speaking, the anthers in front of the sepals also come into contact with the stigmas 

 at the same moment, but in a few cases they pi'oject above the stigmas and petals, 

 and their pollen is then not available for autogamy. It is remarkable that in the 

 the latter, which may be represented by Sagina saxatilis, the characteristic fact 

 of the pollen of the five stamens opposite the sepals being devoted to cross-fertiliza- 

 tion, and that of the five stamens opposite the petals to autogamy, is exactly the 

 reverse of the arrangement found to exist in the Saxifrages above desci'ibed. 



Next to this series of plants of which the Saxifrages of the Aizoonia and Tridac- 

 tylites groups and the Alsinete above-named are the chief repi'esentatives, comes 

 another composed p)i-edominantly of Crueiferje. They are for the most part annual 

 species with small flowers, which are but little visited by insects, and the majority 

 of their fruits must be looked upon as products of autogamy. Cochlearia Gi'een- 

 landica, Draba borealis, Braba verna, Clypeola Messanensis, Lobidaria nummu- 

 laria, HutcJiinsia alpina, SchievereJcia Podolica, Lepidium Braba, Alyssum 

 calycinum, are a few examples, and the selection shows incidentally that the i-ange 

 of the cruciferous plants in question extends from the extreme North to the latitude 

 of the Sahara, and from high altitudes to the level of the deep-lying steppes; in fine, 

 that this same process of autogamy recurs under the most diverse external condi- 

 tions. All these Cruciferas are protogynous, and have six stiiT stamens, four long 

 and two short. The anthers of the former are still closed when the flower opens, 

 but are already on the same level as the stigma. Autogamy is, however, prevented 

 immediately on the dehiscence of the anthers, owing to there being a little horizontal 

 interval between them and the stigma. It is not till the flower is almost over that 

 the erect filaments move sufficiently towards the middle of the flower to deposit the 

 pollen upon the stigma. The pollen of the shorter stamens does not get transferred 

 at all to the stigma in the same flower except in a few species. It is carried away 

 by insects and used for cross-fertilization, whilst the pollen of the longer stamens 

 mainly subserves the purpose of autogamy. Lepidium Braba exhibits a curious 

 contrivance to prevent the four longer stamens from being touched by insects and 

 despoiled of their pollen during the first part of the time that the flower is open. 

 The stamens referred to bend outwards and hide themselves for a time behind 

 the petals. The advantage of this movement is that in no circumstances can there 

 be a deficiency of pollen for the ultimate process of autogamy. In Htdchinsia 

 alpina usually only one of the four longer stamens appi-oaches sufiiciently near to 

 the stigma to cover it with pollen, and after it has effected this object, it removes 



