BY E. HAVILAND. 185 



the base, but connected at the top ; the anthers forming a com- 

 paritively long fluted cylinder, so closely adnate, that considerable 

 force with the dissecting needle is needed to separate them. In 

 fact I have rarely succeeded in doing so cleanly, one anther 

 generally tearing away part of its neighbour rather than separate 

 from it. No trace of the apex of the pistol can be seen ; it is, as 

 yet, so completely closed in at the bottom of this anther tube. 

 Choosing a more advanced specimen, pollen may be seen just 

 beginning to emerge from the top of the tube. In one still more 

 advanced the pollen is seen crowded out of the tube and falling 

 over its side. In another the point of the style can be seen 

 emerging from the tube, and following the pollen which it has 

 pushed before it. When the style has so far advanced that its 

 apex is quite clear of the anther tube, it will be observed that it 

 is crested with short stiff hairs or bristles, which there is no doubt 

 have been used as a brush to sweep the pollen before it. I have 

 not, myself, as a rule, found the style retaining any of the pollen, 

 except occasionally a grain or two. It appears to accumulate on, 

 and cling for a short time, to the outside of the anther tube, 

 allowing the apex of the style to pass beyond it. Examining 

 other and still more mature flowers ; the style will be found 

 projected to different distances, from one to three-eights of an inch 

 beyond the anthers, but no trace of a stigma can be found, and 

 it will be noticed by this time, that in almost every case the 

 pollen has entirely disappeared. It is now that the apex of the 

 style splits into two rather broad spreading lobes, and it will be 

 seen that the inner surfaces of these lobes form the stigmas, 

 which having been enclosed within the style, till all the pollen from 

 their own flower had disappeared, are at last exposed to the visits of 

 pollen bearing insects. As, however, a small portion of pollen is 

 generally left inside the anther tube after the style has escaped, I 

 thought it not unlikely that, at a subsequent period, and after the 

 lobes of the style had opened exposing the stigmas, the anthers 

 would also separate exposing the residue of the pollen. In such 

 case an insect would, in passing from the anthers to the stigmas, 

 undoubtedly self-fertilize the flower. I have, however, found no 



