STRUCTURE.] ANOMALOUS DEHISCENCE. 5 



where the letters have the same value as above. In this 

 it was formerly said that the dissepiments were opposite the 

 valves. 



In septifragal dehiscence the dissepiments adhere to the 

 axis and separate from the valves, as in Convolvulus ; or in 

 the diagram 171., lettered as before. 



In sutural dehiscence there are no dissepiments, the fruit 

 being composed of only one carpel, as the Pea. 



Besides these regular forms of valvular dehiscence, there is 

 a very anomalous mode which occurs in a very few plants, and 

 is called circumscissile. This takes place by a transverse cir- 

 cular separation, as in Anagallis ; in Jeffersonia it only takes 

 place half round the fruit. In some cases, as in lomentaceous 

 legumes, the transverse disarticulation may be supposed to 

 have some relation to the pinnate leaves, whose modification, 

 in those instances, forms the carpel. In other cases the 

 explanation is far less obvious, and must be at least very 

 different. Perhaps the best account of transverse dehiscence 

 is that of Mr. Hincks, as reported in the Annals of Natural 

 History, vol. xvii. 



" In the fruit, as in the calyx, this author believes that 

 horizontal disruption arises from the force of cohesion of the 

 parts of the circle, the absence of any of the causes favourable 

 to dehiscence along the midrib of the carpellary leaf, and the 

 operation of some force pressing either from without or from 

 within on one particular line encircling the fruit. In the 

 circumscissile capsule of Anagallis, he states that the central 

 free receptacle with the seeds upon it continuing to enlarge 

 in both diameters after the envelope has ceased to grow, 

 and having occupied from the first the entire cavity, it is 



