POPULAR SCIENCE REVIEW. 
2(i8 
I come now to a number of species in which the lower anther 
cells are still further modified, being transformed into little 
hollow plates. The concavities of these plates are turned 
upwards, and by the union of the plates a little boat-shaped 
receptacle is formed, which more or less completely blocks the 
opening into the tube. (See figs. 9, 10, 11.) Under this head- 
iug come S. sclarea, jpmtensis, sylvestris, grandiflora, patens, 
splendens, and a vast number of others. The most interesting 
of them is S. patens. This one, therefore, 1 will describe. 
S. patens. This large garden-flower can be easily obtained 
and examined. It will be found that there is a different 
mechanism for its fertilisation from that of any other species I 
have mentioned. The style, when it reaches the lower part of 
the hood, passes from behind forward between the two anthers, 
and higher up passes back again between them a second time, and 
then projects above them from the lip of the corolla. (Fig. 12.) 
At the points where it passes between the anthers, it is held firmly 
l3y them ; and thus, when the anthers rotate the style moves 
with them, coming forwards and retiring back in company with 
the upper anther cells. These and the stigma retain their rela- 
tive positions to each other while this motion goes on. (Fig. 13.) 
Should now a large insect visit the flower and push the lower peta- 
loid anther cells in order to get at the nectary, its back will be 
struck not only by the polliniferous anther cells, but also a little 
farther back by the stigma. As the insect passes deeper into 
the tube, the anthers and stigma Avill be rubbed along its back 
in a direction from before backwards, the stigma being always 
in advance of the anther cells, and therefore not collecting any 
of the pollen. As the insect retires from the flower the anthers 
and stigma retreat into the hood. The insect flies off to another 
flower, and now the motion brings down the stigma on to the 
pollen-smeared place in its back. It will be noticed that, in this 
species, the upper division of the stigma is very much larger 
than the lower, which is in fact almost abortive, whereas in most 
<)tlier species the lower division is the lustier. The possible 
piirport of this is to cause a greater interval between the pollen 
and the stigma, and thus to render the chances of self-fertilisation 
smaller. On examining the corolla tube, it will be noticed that 
there is a curious constriction in its lower part. By this the 
calibre of the tube is reduced at that point to a very small pas- 
siige, and this passage is filled up completely by the style which 
runs through it; so that, in fact, at this point the tube is entirely 
blocked up. On now examining the inner surface of the tube 
with a microscope, it will be seen that the part above the con- 
striction is thickly set with glandular hairs, while the part below 
is entirely devoid of them. It is by these glandular hairs that 
the fluid which attracts insects is secreted, and if a tube be 
