12i JOURNAL OF THE EOYAL HORTICULTURAL SOCIETY. 



insect till they remained there, so that now the four anthers stand 

 together under the "hood" of the corolla. The style is bent backwards 

 in the same way ; and then, arching forwards, the stigma is brought to 

 the front ; and the whole flower is thus in perfect adaptation to the bee, 

 Avhich thrusts its thorax against the anthers of the stamens above it, 

 while the stigma strikes it on the same spot. Such is the method of pol- 

 lination in the Deadnettle. 



In Salvias the stamens have become even more curiously adapted, two 

 only being effectual. They have very short filaments, and the anthers 

 swing up and down upon them, as on pivots. The anthers are extra- 

 ordinarily modified. By reference to fig. 52 it will be seen that the 

 connective which joins the two anther cells is elongated into a curved rod, 

 carrying one anther containing pollen at the top, while the other, which is 

 open and devoid of pollen, is at the bottom. They constitute two levers, 

 so that when a bee's head depresses the spoon-like lower extremity, the 

 upper anther cells swing down into the position shown in fig. 52, and 

 strike the bee on the back, as shown in fig. 51. On entering another 



flower, the stigma hits the bee where the pollen had been previously 

 deposited. 



Now, where are the honey-glands in these irregular flowers ? We 

 have seen that in regular flowers, as the Geranium and Raspberry, the 

 supply is obtained from five glands situated at regulari intervals on the 

 floral receptacle in the former, and in a circular trough in the latter, 

 because the insects can visit the flowers from any quarter. Itiis otherwise 

 with irregular flowers. The honey is accordingly supplied from one gland 

 only, and this is situated precisely where the proboscis of the insect can 

 get it. Sometimes it is on the front or anterior side, as' in Deadnettles, 

 &c. ; sometimes on the back or posterior side ; but in each case it is 

 exactly in keeping with all the other details of the flower. 



Besides the above peculiarities special maikings are upon the petals, 

 called "guides," or "pathfinders," to call the insect's attention to the 

 right direction where to thrust its proboscis. If the gland is on the 

 anterior side, then the markings are on the lip petal immediately over it, 

 as in the Violet and Pansy, Labiates, &c. But if it be on the opposite side, 

 then the guides will be found on the posterior petal or petals, as in Pelar- 

 gonium, Rhododendron, &c. 



We thus see what an accumulation of features there are, all conspiring 

 to one and the same end — namely, the adaptation of the insect to obtain 



a 6 



Fig. 55. — a. Flower of Diivernoia 



adhatodoides ; b, same visited by 

 humble-bee. 



