350 NATURAL SCIENCE. may. 



period ; the flowers developed under these conditions being carefully 

 measured and compared with the normal type. The first plant 

 described is Miinulus Tilingii ; here the flowers, developed in poor light, 

 were smaller than the normal, and in the extreme cases did not 

 develop at all, but aborted in the bud stage. An interesting point m 

 the gradual diminution of the corolla was the difference in the rate of 

 decrease of the upper and lower lips. Down to a certain size of 

 flow-er their proportionate dimensions were nearly constant, but below 

 this the upper lip decreased much more rapidly than the lower. In 

 Linaria spuria, a plant which very commonly bears cleistogamic 

 flowers in the wild state, the decrease of illumination resulted in 

 flowers of this type only. Similarly with Laminm amplexicaiile and 

 other plants. Chickweed {Stellaria media), it was found, could be 

 made to produce either ordinary or cleistogamic flowers by suitable 

 illumination. In the wild state it is cleistogamic in bad weather. 

 In Melandrium album it was observed that the corolla was the first 

 organ to show a decrease with diminution of the light. From these 

 experiments it follows that, for the normal development of a flower, 

 a certain intensity of light is required ; but this varies largely in 

 different species. Low intensity of light results in reduced flowers, 

 the reduction showing first of all in the corolla ; in plants with a 

 tendency to cleistogamy, flowers of this kind are produced. Hence 

 we may look upon decrease of light, perhaps, as the primary cause of 

 cleistogamy, calling forth incipient flowers of this type, upon which 

 natural selection may come into play. Vochting also observed, most 

 markedly in Mimuhis Tilingii, that the low degree of illumination tended 

 to increase the vegetative activity of the plant. In Mimuhis the axis 

 of the inflorescence gave rise, instead of flowers, to numerous vegeta- 

 tive shoots. By suitable treatment, the flowers were prevented even 

 from beginning development, and vegetative reproduction alone 

 occurred. Cleistogamy has also been studied by Graebner, who 

 considers that it may be brought about by other agents as well, 

 e.g., by temperature or the weakening action of fungi, etc., or, again, 

 by an unsuitable relation of light to temperature (if one be raised, the 

 other must also be raised, or abnormal floral development will occur). 

 This view is also taken by the present writer, who has studied the 

 behaviour of Salvia verbenacea and other plants. 



It is thus evident that a study of the natural history of the flower 

 from a physiological standpoint has already led to important results 

 and may be expected to yield many more in the future. The 

 present Avriter has studied the phenomena bearing on the various 

 distributions of sex by the experimental method during the last 

 four years, drawing therefrom the general conclusion that all these, 

 as well as cleistogamy, are primarily due to the action upon the 

 plant of external causes. Natural Selection apparently only begins 

 to act later on, when the phenomenon in question has been to some 

 extent produced by these causes. 



