394 AUTOGAMY. 



underground, their stalks springing from special shoots of the rootstock. In all 

 these eases the two kinds of tlowers are always borne on the same plant, though on 

 different branch systems ; there are, however, also instances, such as the Yellow 

 Balsam (Impatiens Noli-tangere), where the open iiowers are developed on different 

 individuals from those which produce the closed ones. To be accurate we should 

 say that the statement in each case expre.sses the genei-al rule for the plants in 

 question, for instances of transition are by no means uncommon. Thus, for example, 

 individual plants of the Yellow Balsam do occur in which open flowers with large 

 corollas, half-open flowers with stunted corollas, and small cleistogamous flowers 

 stand side by side; and, again, on the zigzag runners of the Sand Violet {Viola 

 arenaria) flowers with large expanded petals have often been seen gi'owing in 

 company with the cleistogamous flowers. The same remark applies as regards the 

 time at which cleistogamous flowers make their appearance. In the majority of 

 cases they are not developed until the open flowers have withered and disappeared, 

 but in Cardamine chenopodiifolia it has been observed that the subterranean 

 cleistogamous flowers are produced earlier than those which are borne on above- 

 ground stems and unfold their petals to the air. 



In former times it was asserted that plants exist which never bear any but 

 cleistogamous flowei-s. Thus the Toad -rush {Junctcs bufonius) was stated to 

 produce cleistogamous flowers only. Later investigations have, however, established 

 the fact that this plant possesses two kinds of flowers — viz. terminal flowers with 

 three stamens which are cleistogamous, and lateral flowers with six stamens which 

 open, like those of other Rushes, under the warmth of the noonday sun. It was 

 also supposed that the African species of Scdvia, to which the name of Salvia 

 cleistogama was given, produced cleistogamous flowers only, but after repeated 

 sowings plants have been obtained with flowers which opened. Anyone who has 

 only seen the Yellow Balsam growing on the heaps of detritus brought down by 

 the mountain-streams in the upland valleys of the Tyrol might think that this 

 plant also was an instance of a species f)roducing cleistogamous flowers only; for in 

 those localities open flowers are not found on this species. But if seeds from these 

 cleistogamous flowers are sown in good vegetable mould, in a partially shaded spot 

 in a garden, a few individuals with large open yellow flowei-s spring up, as a general 

 rule, even after a single sowing. There is a species of Violet named Viola sepincola 

 which grows deep in the shade of the woods clothing the hills at the foot of the 

 Solstein chain in the Innthal district. I saw it there for the first time about the 

 middle of May, and it was then covered with an abundance of ripe fruit. In follow- 

 ing years I looked for flowers of this plant early in the spring, as soon as the snow 

 had melted, but found that not a single individual had developed open flowers with 

 expanded petals on erect above-ground stalks. On the other hand, there were a 

 number of cleistogamous flowers concealed under the fallen leaves and partially 

 buried in the earth, so that it looked veiy much as if the species produced no other 

 kind of blossom. But plants subsequently reared in a part of my garden which was 

 exposed to the sun's rays during some hours of each day developed, in the next 



