MR. J. MIERS ON THE FAMILY OF TRIURIACEjE. 53 



geneous, which would not accord with its nature ; but if these denote only different con- 

 ditions of the embryonary form, the one may be considered as a protoblastous, the other 

 as a cotyledonous embryo. On surveying other peculiar embryonal forms, some will be 

 better understood by this view of the subject ; for instance, we may conceive that the pro- 

 toblastus, instead of forming one compact mass of spherules, as in Balanophorece, Trm- 

 riacece, &c, may be somewhat less aggregated, so as to assume the shape of elongated 

 bundles of cells, such as have been termed paraphysiform : such a form is actually met 

 with in Ceratophyllum, where the exterior series of unequal size have been assumed to be 

 an unusual number of cotyledons, and the very numerous inner series have been held to 

 be a highly developed plumule, while the common point of their union is considered 

 as the radicle. The development of this embryo has been well analysed by Scldeiden, 

 in his memoir on Ceratophyllum *. Professor Lindley remarks t, that " in this instance, 

 as in Nelumbiacece, the highly developed plumule may be a compensation for the want 

 of albumen, enabling the embryo to germinate without assistance, as soon as it is ex- 

 posed to the fitting conditions." The leaves of Ceratophyllum appear destitute of all 

 nervures, consist of confervoid parallel cells, and dichotomously divide themselves into 

 simply articulated hair-like segments, thus denoting a lower order of development than 

 has been assigned to it. Independently of this cellular texture, we must bear in mind 

 that this genus possesses monoecious flowers, with a simple perianth, having a valvate 

 aestivation, almost sessile stamens, an unilocular carpel with a solitary ovule ; and if, 

 in addition to these characters, we regard the construction of its embryo, in the view 

 above contemplated, its position in the system would rank near Fluviales, as was long 

 ago suggested by Bernard and Antoine de Jussieu, followed by Jaume St. Hilaire and 

 Agardh, and therefore, as will be seen, not far from Triuriacece. Conterminous with the 

 ■Fluviales we find another family, the Aroidea?, which offers many circumstances bearing 

 upon this subject. Blume describes Amorphophallus% as possessing an exalbuminous 

 simple nucleus, homogeneous in texture, with one of its extremities pullulating at one, 

 two, or three points, and throwing out fleshy lobes, which overlap each other. The same 

 botanist records, that in Aglaonema the seed has a solid nucleus, which, in germination, 

 throws out several squamulse at one end § ; and Schott states that the nucleus of Crypto- 

 coryne emits several gemmulae in a similar manner ||. This fact has been confirmed 

 and illustrated in an admirable manner by Mr. Griffith, in a very interesting 

 memoir upon Ambrosinia (Cryptocoryne) ciliata%, where he has shown, that at an early 

 period the ovule presents an embryo, which then appears to be quite homogeneous, and 

 " entirely cellular," and that its development is first marked by the production, upon a 



* Linneea, vol. ii. p. 512. f Vegetable Kingdom, p. 263. 



t In Rumphia, i. 138 : "in quo ad extremitatem inferiorem umbilicum spectantem una rarius 2 vel 3 minutse gem- 

 mulae germinantes observantur ; gemmulae ejusmodi constant squamulis aliquot carnosis sibi oppositis et sese amplec- 

 tentibus." (Kunth. Enum. iii. 32.) 



§ "Embryo semini conformis, exalbuminosus, solidus, ad extremitatem radicularem squamulis aliquot munitus." 

 — Rumph. i. 130. (In Endl. Gen. PI. Suppl. p. 1370.) 



II "Embryo cotyledonibus (protophyllis ?) plurimis."— Schott, Meletem. Bot. (In Kunth. Enum. iii. 12 ; et Trans. 

 Linn. Soc. vol. xx. p. 266.) 



^f Trans. Linn. Soc. vol. xx. p. 263, 



