2i6 Alexander W. Evans, 



ber, but the third row commonly shows twice as many and the 

 fourth row a much larger number. The ridge immediately around 

 the opening is clearly marked. In the lower series (Fig. 19, H-K) 

 the innermost row lies almost directly beneath the second row, 

 so that only the first and third rows show clearly from below. 

 The first and second rows are usually composed of four cells 

 each, the walls bounding tlie pore being more or less strongly con- 

 vex. The third row usually contains more cells than the first and 

 sometimes twice as many, but it rarely contains as many as the 

 fourth row of the tipper series,' where the cells are essentially 

 like the ordinary epidermal cells. Of course the numbers just 

 given are subject to variation, the number of cells bounding the 

 outer and inner openings being often more than four. 



As an example of the second type of pore M. paleacea may be 

 selected, and the illustrations given in the present paper (Fig. 

 8, A-H) may be compared with the one published by ]\Iuller.-^ 

 The descriptions given by Voigt may likewise be consulted. The 

 cells bounding the pore are usually in six rows, three belonging to 

 each series, and the rows are commonly composed of four cells 

 apiece. The cell-walls bounding the pore are smooth through- 

 out. In the upper series the ridge around the opening is distinct 

 and the cells are very narrow, standing in sharp contrast to the 

 neighboring epidermal cells. In the lower series the cells bound- 

 ing the inner opening are much broader than the others and project 

 so strongly that they often touch in the center and almost occlude 

 the cruciate opening. Sometimes one or more cells of the second 

 row project also (Fig. 8, D), but the cells of both the second 

 and third rows are usually narrow, resembling in this respect the 

 cells in the upper series. 



According to the account given by Kamerling,-* the size of the 

 inner opening in a pore of the first type is not decreased to any 

 great extent by an increase of the turgidity of the surrounding 

 cells, while in a pore of the second type the decrease is very 

 marked. In his opinion the cells surrounding the inner opening 

 act independently of the cells in the other rings. As a definite 

 example of a species with pores that can be closed he quotes M. 

 nitida Lehm. & Lindenb., a species which is to be regarded as a 

 synonym of M. paleacea. 



" Rabenhorst's Kryptogamen-Flora 6 : /. 188. 1907. 

 -^ Flora 84 (Erganzangsb.) :46. 1897. 



