364 FERTILISATION OF THE GOODENIACE.E, 



against the fold of the anterior margin and so moves the auricles 

 apart, when the pollen falls in a small shower on its thorax and 

 head. All these contrivances point towards insect-fertilisation, 

 but to complete the process one would imagine that the stigma 

 should now grow beyond the lips of the indusium and project as 

 has been described in SccEvola and Selliera [5]. But in at least 

 ninety-five per cent, of the flowers the stigma does not project at 

 all beyond the mouth, and in many it does not grow up to the 

 level. Examination of large number's of flowers just withering 

 showed the outside of the indusium, the stigma, and between the 

 stigma and the inner side of the walls of the indusium coated 

 evenly with pollen all over, and this was apparently caused by the 

 close fit of the auricles round the style applying the pollen, and 

 not by insect agency at all. I can onl}'- conjecture that the plant 

 is ordinarily self-fertilised, although occasionally cross-fertilisation 

 may occur from insect visits when the stigma is outgrown or near 

 the mouth of the indusium. Erom the firmness of the hold whicli 

 the auricles have upon the style, any insect would, in forcing its 

 way in, press hard against the style and might thus deposit pollen 

 upon the stigma, even though it was only at the mouth of the 

 indusium, instead of projecting as in other genera. It is usual to 

 find the auricles full of pollen where the flower and even the 

 style is withered, so that insects do not commonly exhaust the 

 pollen. The flowers are very sweetly scented, and there is a 

 considerable amount of moisture at the base of the petals, in 

 which, however, I could not detect any sweetness. As already 

 pointed out, the membranous edges of the three lower petals are 

 closely pressed together in the tube, and pucker so as to form 

 guiding ridges (Fig. 3). The flowers are much frequented by 

 Thrips. The auricles in the early bud are green, without any 

 deep concavity, and with a pale red spot where the deepest colour 

 occurs in the mature flower; this is indicated by the dotted oval 

 in Fig. 4, which represents a young auricle. In this stage it 

 resembles tlie mature auricle in D. linearis (Fig. 12). It gradually 

 darkens till it is a fine purple-red with the central part a purple- 

 black of wonderful intensity. This bears out Dr. A. R. Wallace's 



