1352 A TEXTBOOK OF THEORETICAL BOTANY 



tricha. In Britain we have the case of Triodia decumbem, and it seems most 

 probable that further search in the Gramineae would reveal many more. 



In some of the above-mentioned examples, cleistogamy is more or less 

 constitutional, but in others, and in a great number of tropical and sub- 

 tropical species, it is ecological, that is to say that the presence or absence of 

 cleistogenes is controlled by environmental factors. Among such factors 

 are : submersion in water {Elatine and some Ranunculus species) ; prevalence 

 of mist and high atmospheric moisture {Liparis in New Guinea); drought, 

 in the case of species normally demanding high soil moisture {Impatiens 

 parviflora) ; cold {Specularia and some montane species of Viola and Thlaspi); 

 heat, if excessive {Eranthemum); shade {Viola arvensis, V. sepincola and 

 Linaria vulgaris). Briefly the observations sum up to indicate factors un- 

 favourable to the full development of the species in question, but not 

 all species respond to such influences by cleistogamy. Unless there is a 

 hereditable tendency to cleistogamy the majority of species merely fail to 

 flower at all when conditions are bad. 



The origin of cleistogamy is uncertain. According to Goebel's view, 

 cleistogamous flowers correspond in structure to normal (chasmogamous) 

 flowers of inhibited development, but endowed with the power of preco- 

 cious pollination and therefore capable, in spite of their imperfect develop- 

 ment, of forming seeds. This is contrary to Darwin's opinion, who held that 

 cleistogamous flowers showed special modifications to ensure self-pollina- 

 tion. It is clear that they are not always ecologically conditioned, since they 

 may be formed alongside of normal flowers on the same plants. Engler's 

 observations on Streptocarpus led him to conclude that the inhibition of 

 development in cleistogenes was the effect, not the cause, of their preco- 

 cious self-pollination, which he attributed to inner causes which brought 

 pollen and stigmas to functional maturity together at an early stage. In 

 West African species of Streptocarpus, belonging to the section Caules- 

 centes (Fig. 1260), some species produced only cleistogamous flowers, 

 others had one or two chasmogamous but sterile flowers, as well. He further 

 showed that in the many species of the genus which bear only chasmogamous 

 flowers, self-pollination might occur. 



Goebel's theory might apply in some cases where cleistogamy appears at 

 a certain stage in the life history of the plant, but it can hardly apply to a 

 species like Salvia cleistogama, where cleistogamy is a fixed specific character 

 and no chasmogamous flowers are produced. In some South American 

 species of Plantago cleistogamy shows a sex-linked inheritance. Some plants 

 have only hermaphrodite, cleistogamous flowers, while other plants are 

 wholly male. The hermaphrodite flowers, cleistogamously pollinated, give 

 only cleistogamous offspring but if pollinated from male flowers the off- 

 spring are all male. Variations of nutrition had no effect on the condition. 



Mather and Vines found two cleistogene plants among the offspring of a 

 cross between Antirrhinum majus and A. glutinosum. These plants had 

 distinct foliage and set good seed, the cleistogamy being heritable. Thus 

 two species, one an obligatory cross-breeder and the other a facultative 



