MR. J. MIERS ON THE FAMILY OF TRIURIACE^. 55 



be compared with the arilliform covering above described in the genus just mentioned. 

 The embryo of Pistia is cylindrical, nearly half the length of the albumen, in the summit 

 and in a longitudinal cavity of which it is placed, with its apex quite bare, the whole being 

 enveloped by a thin pellicular integument, and this again by a thin and somewhat coria- 

 ceous reddish testa, marked much after the manner of that of Soridium. This testa is 

 fixed upon a long, thickened, stipitate support, and is crowned at its apex by a dark 

 pulvinate process, suspended by a cylindrical plug or strophiole, the point of which is seen 

 in the umbilical apex of the seed. The external covering is a thick, wrinkled, fungous or 

 pithy substance, lined inside and outside by a thin adhering membrane, and enclosing the 

 stipes, the testa and the strophiole, and it can be considered in no other light than an arillus, 

 and as analogous to the peculiar covering of the testa seen in Sciaphila. In the Zingibera- 

 cece the seed is also covered by an arillus somewhat similar to that of Pistia, while in Rave- 

 nala, by its numerous ribs, it approaches yet nearer to the seminal envelope of Sciaphila ; 

 here, however, as well as in Pistiece, the albumen is of peculiar structure, being farinaceous 

 when dry but becoming distinctly cellular when moistened, and the cells are separable 

 from each other without bursting ; they are then translucent, and bear altogether a very 

 different aspect from those seen in the nucleus of Sciaphila, in which they are more 

 opake, as if filled with dark grumous matter. The facts here brought together are inter- 

 esting, as presenting some analogies bearing upon the question, and they serve to show 

 that we have yet much to learn concerning the nature of the more simple forms of em- 

 bryonal structure, and of the functions of reproduction. 



In respect to the position of the Triuriacece in the system, if we follow the rules of 

 classification founded upon the three great divisions of Acotyledonous, Monocotyledonous, 

 and Dicotyledonous plants, a plan now quite untenable, they must of course range in the 

 same group as the Balanophorece, although they hold but little relationship with them. But 

 if we regard the condition of the embryo, not as a basis, but merely as a frequent indication 

 of the three great divisions, founded on the structure and development of vegetable fibre, 

 viz. Acrogens, Endogens and Exogens, and if we accord with the views of Mr. Brown and 

 Mr. Griffith, in considering what the latter calls the homogeneous-embryo-form state of 

 the seed merely as an imperfect condition of development, common alike to all these 

 groups, then the Triuriacece must take their place among the Endogens. 



Here, the family that at first sight appears most approximate is the Alismac'ece, with 

 which the Triuriacece agree, in their simple stem, sometimes spicate unisexual flowers, 

 and their numerous carpels ; but Alismacece differ in having leaves with parallel nerves, 

 in their floral envelope being distinctly biserial, the outer calycine, the inner petaloid, and 

 both with imbricated aestivation ; they are also dissimilar in their bilocular anthers, with 

 parallel cells, dorsally affixed to long filaments ; ovules often two in each carpel, one being 

 superimposed ; carpidia opening by their ventral suture, and seeds with large hippocrepi- 

 form embryo. 



With Fluviales, Triuriacece accord in their soft cellular structure, their monoecious 

 flowers, simple perianthium with valvate aestivation, often 4-locular anthers, several 

 distinct ovaria, with a single erect ovule in Caulinia and Najas, baccate fruit, with a 

 pericarpial utricle bursting on one side in Althenia, Zoster a, Najas and Cymodocea ; but 



