Evans : Hepaticae of Puerto Rico 11 



times the proliferation of an antheridial spike, sometimes the inno- 

 vation of a female inflorescence. In any case the growth of the 

 branch is sooner or later brought to an end, although in one ob- 

 served instance as many as fifty modified leaves had been developed. 

 The formation of these peculiar branches is apparently induced by 

 crowding*, and it is not unusual to find them in the middle region 

 of a tuft, the marginal part of which continues to develop branches 

 of the normal type. 



The modified leaves are strikingly different from ordinary 

 leaves in their appearance. They are densely imbricated, and the 

 line of insertion is nearly transverse and very short, measuring but 

 five to ten cells in length. No lobules are developed, and the 

 lobes are ovate in shape and almost symmetrical. The margin is 

 not entire, as on normal leaves, but bears a series of scattered hair- 

 like teeth, which vary in number from one to ten. The teeth are 

 irregular in arrangement, but the one at the apex is usually dis- 

 tinct, thus making the leaves acuminate. Each tooth is composed 

 of a single row of cells or is rarely two cells wide at the very 

 base ; it is sometimes straight and sometimes variously curved or 

 hooked at the tip. In the majority of cases the terminal cell is 

 long and delicate, like a rhizoid, and occasionally the entire tooth 

 is reduced to a cell of this character. The leaf-cells have much 

 thinner walls than ordinary leaves and are well supplied with 



chloroplasts. 



The leaves just described develop rapidly and become detached 

 as soon as they attain their full size. The line of rupture is very 

 close to the base, and the detached leaf leaves behind a narrow 

 ridge of projecting cells, the process of separation being schizolytic 

 in nature. A separated leaf is able to live a considerable time on 

 account of its many chloroplasts, but it soon gives rise to one or 

 more leafy shoots, without the interpolation of thalloid structures. 

 Apparently any marginal cell has the power of developing into a 

 shoot of this nature (figure 24), but the number growing from a 

 single leaf is always very small. 



The underleaves on the flagelliform branches are not decidu- 

 ous ; they are closely imbricated like the leaves, and are more or 

 less squarrose, thus giving the branches a peculiar and characteristic 

 appearance. The underleaves do, however, exhibit marked modi- 



