402 



BOTANICAL GAZETTE 



[DECEMBER 



whether it could possibly be homologous with an air chamber. 

 From an a-priori consideration of the general character of the cupule 

 and the air chamber this idea commended itself to us both. The 

 walls of the cup, especially the thin, lobed margin, seemed to corre- 

 spond very well with the epidermal roof of the air chamber, opened 

 wide instead of having only a narrow orifice. The gemmae, borne 

 upon a single cell arising from the floor of the cup, might well be 



Fig. i. — Early stage of cupule; p, p', 

 undivided cells, primordia of two ( ?) 

 gemmiparous areas; a, apical cell; air 

 chambers shaded. 



Fig. 2. — Early stage of cupule; p, p', 

 as in Jig. i ; w, probably a rim cell 

 between two gemmiparous areas ; a, 

 apical cell; /, /, line showing tissues of 

 thallus involved in a cupule. 



only a modified form of the chlorophyllose filaments of the air cham- 

 ber. So natural and neat did the homology appear, that the brief 

 prior statement of it by Campbell (I. c.) was discovered with a dis- 

 tinct sense of disappointment when we began to look into the litera- 

 ture. But evidence for this homology could not be found therein, 

 and against it was to be put the fewness of the cupules, their limita- 

 tion to the median line, where the air chambers are least developed, 

 and the fact that the gemmiparous region covers many times the 

 area of an air chamber. The matter evidently needed examination. 

 Actual observation of the origin of the cupule speedily dissipated all 



