ENGELMANX — THE GENUS ISOETES IN N- AMERICA. 365 



no fresh spores are attainable. The root-fibres, sometimes longer 

 than the leaves, are always dichotomously, and often many times^ 

 branched. — The upper, concave, surface of the trunk bears the 

 leaves, the innermost or youngest ones often yet immersed in 

 the trunk. 



The LEAVES are subulate or sometimes almost filiform tubular 

 organs from a broad membranaceous sheathing base, mostly more 

 or less quadrangular (broader and with sharper edges on the up- 

 per or ventral, narrower on the dorsal side), or in our terrestrial 

 species more triangular and keeled on the back. Their sheath- 

 ing bases form the hulb^ which can be compared to the bulb of 

 liliaceous plants; in fertile specimens it is always larger and 

 thicker than the trunk, and in some of the larger ones, e.g. /. 

 Engelmanni diVid. /. melanopoda^ attains a diameter of one or two 

 inches. The leaves above this base contain four longitudinal air 

 cavities, lacunce^ separated from one another by two dissepi- 

 ments, a transverse and a median one, and irregularly divided by- 

 very thin transverse septa. The dissepiments are of different, 

 pretty constant, thickness in the different species, thinnest in the 

 amphibious and thickest in the terrestrial species, consisting in 

 the former often of only 2 to 4, in the latter of 6 to 9 layers of 

 parenchymatous cells; the median dissepiment is generally a 

 little thicker than the transverse one. The anterior lacunce are 

 mostly somewhat larger than the posterior ones. 



The epidermis of the leaf consists of rectangular cells, mostly 

 much longer than they are wide ; only in /. pygmcsa are they 

 comparatively short, and sometimes even square. In a few spe- 

 cies the epidermis is entirely destitute of stomata, in the others it is 

 pierced by stomata which communicate with the air-ducts, over 

 which alone they are found. The presence or absence of stomata 

 furnishes a very important character for the diagnosis and classi- 

 fication of the species. It was formerly thought that the sub- 

 merged species had no stomata, and those species which bear 

 their leaves more or less exposed to the air were provided with 

 them ; later discoveries, however, have_ shown that this rule does 

 not always hold good, for we now know submerged species with 

 stomata and emerged ones without them, and we have one sub- 

 merged species (/. echinospora) in which the typical European 

 form is destitute of stomata, while the American varieties show 



