Francis E. Lloyd — 256 — Carnivorous Plants 



more and more concave till the set posture is reached. This account 

 has been substantiated by a photographic record of silhouette of living 

 traps in the set and relaxed condition of two related species. Seeds of 

 U. Wehvitschii collected by Young in Angola were grown for me at 

 the Edinburgh Royal Botanical Garden in 1934 and the traps studied 

 there. U. capensis was studied alive at Capetown later, and the results 

 were pubHshed in 1936 {24 — 5) (Lloyd 1936&). 



Another, an Australian species, U. lateriflora, typical of a small 

 group of species distributed in S. E. Asia, and Australia, having very 

 small traps less than i mm. long, yielded to experimental methods 

 (1936c) and the results are shown in 33 — 9, demonstrating that the 

 behavior is quite like that in U. capensis and U. Welwitschii. The 

 living material was available at Sydney, N. S. W. 



The same behavior is displayed by U. caerulea (Asia) {24 — 2) and 

 by U. cornuta (N. Amer.) (jo — 3) in both of which the living trap 

 was studied. 



In all these, when the trap is in the set condition the outer selvage 

 of the door rests on the pavement, held there firmly by the thrust of 

 the lateral hinges. The wide angle between door and threshold is 

 filled by the massive velum, preventing inleakage of water. A thrust 

 on the tripping mechanism, the kriss hair (p. 259) in U. capensis and U. 

 Welwitschii, a group of sessile glands in U. caerulea and U . cornuta, 

 disturbs a dehcate balance of forces in unstable equilibrium, and the 

 trap is ''fired." 



Both U. monanthos {24 — 3) and Polypompholyx {24 — 8) act in 

 the same way, and they also have been studied in the living condition. 

 U. monanthos was grown for me in Edinburgh in 1934 (1936a) and 

 Polypompholyx could be examined in 1936 at the University of Western 

 Australia at Perth near which it grows. The structures involved are, 

 however, to be considered separately. 



U. monanthos {j4 — 1-5)-- — In this and allied species, the thresh- 

 old is very broad, front to back, and near its inner limit is bent, 

 curving downwards. Beyond the bend lies the dense pavement 

 which receives the middle piece, which is therefore applied on the 

 inside of the bend. This looks like a pretty poor arrangement, yet 

 it works. The major zone in front of the bend is occupied by an 

 ample velum which arises also from the walls projecting in front of 

 the door. Here is formed a complete massive ring resting against the 

 bulge of the upper part of the door when in the relaxed posture. When 

 in the set posture, the inner portion of the velum arising from the 

 pavement alone continues to block the entrance of water. The door is 

 still longer than in U. capensis etc., but the middle piece is relatively 

 smaller, and the middle area is correspondingly large, occupying about 

 four-fifths of the door length. When in the set posture, the whole of 

 this large area is concave, so that the sagittal curve is now continuous 

 with that of the middle piece, which by virtue of the thrust of the 

 lateral hinges is impressed against the dense pavement just inside the 

 bend of the threshold. The trigger consists of a group of sessile hairs 

 just above the bend of the door. The action when the trap is fired is 

 like that in U. capensis. It must be confessed at this point that my 

 earlier account of door action (1932a) based on preserved material of 



