Scutellaria. 117 



moved, magnified 5 times), each of the two lateral petals 

 presents a large bulge at the mouth of the tube. These 

 bulges have their convexities turned towards the centre of 

 the aperture, which they block up to such an extent that 

 a small hole, scarcely 2 mm. in width, is aU that is left 

 between them. Insects, however, in entering into the 

 flower, push these two bulges asunder ; and the conse- 

 quence is that the loose pollen which has been dis- 

 charged from the anthers under the upper lip, and has 

 been deposited on the cushion-like upper surface of the 

 arched bulgings, is displaced from this position, and 

 showered down on the back of the insect. The insect 

 in its further roving soon carries it to some other 

 flower. Supposing these bulges not to exist, no insect 

 would rub against the pollen, unless that part of its 

 body which has to be introduced into the tube were at 

 least 4 mm. in thickness ; but as it is, insects of only 

 half these dimensions can push apart the two bulgings, 

 and in so doing will get besprinkled with pollen, and will 

 therefore be just as useful as larger insects in promot- 

 ing intercrossing. Such is the arrangement ia Scutel- 

 laria. In Ehinanthus and Bartsia the entrance seems 

 to be narrowed in another way, namely, by the three 

 petals that form the lower lip retaining in the flower 

 the same involuted position they had in the bud, 

 and thus barricading the entrance in its lower part. 

 In Calceolaria (Plate II. fig. 52, lateral view of a 

 flower of Calceolaria Pavonii Benth.) the whole lower 



