UTRICULARIA 279 



the visitor, entangled in the bunch of hairs, pushing open 

 the valve, swimming in, the valve falling to behind it, not 

 to be opened by pushing from the inside. But Czaja (1922) 

 has recently confirmed an observation by Brocher (191 1) 

 that the bladder mechanism is active. The walls are im- 

 permeable and water can pass through them only very 

 slowly if at all. On the inner wall are situated numerous 

 four-armed glandular hairs, and these withdraw water from 

 the interior of the bladder. As water is withdrawn the 

 walls are pulled in ; they show a " dimple " on each 

 side instead of being distended. A considerable tension 

 is set up by the mechanical tendency to expand. If now 

 an animal touches one of the hairs on the outside, the 

 valve is levered slightly open ; the tension is relieved, 

 the walls expand, water rushes in, and the animal is drawn 

 in with the current ; the valve closes and escape is 

 impossible. A pricked bladder does not function as no 

 tension is set up, water passing freely in from the outside. 

 Merl (1922) regards the opening of the valve as due to a 

 stimulus ; but Czaja's explanation is more satisfactory. 



No digestive enzyme has been found in Utricularia, 

 perhaps because of the small size of the bladder. It is 

 generally assumed that the captives swim about inside till 

 they die, that they then undergo bacterial decay, and that 

 the products are absorbed, probably by the glandular hairs. 



We have said that the vegetative portion of Utricularia 

 seems to consist of shoots bearing divided leaves, but this 

 interpretation of the morphology is probably incorrect 

 and certainly insufficient, as is shown by comparison with 

 the tropical land species. From the base of the flowering 

 axis of Utricularia Hookeri, a West Australian species, there 

 spring three different types of organs — simple hnear foliage 

 leaves, ^ cm. long, bladders on short stalks, and long 

 rhizoids which penetrate the soil. All three are homo- 

 logous ; in their relation to the parent axis, and in their 

 developmental origin they are identical. The bladders 

 and the rhizoids must be regarded as modified leaves, and 

 this gives an idea of the range of form which may be shown 



