of the Fossil Plants of the genus Sigillaria. 7 



the inner part of the radiated cylinder a festooned appearance, in con- 

 sequence of each bundle forming, as it were, the segment of a circle, 

 having its convexity towards the supposed pith. The vessels composing 

 these bundles are irregularly angular, and arranged without any order, 

 with the exception that the widest, which are also the longest, imme- 

 diately adjoin the pith, while the narrowest and shortest are on the op- 

 posite side : the extremities of both kinds terminate in obtuse cones. 

 By means of longitudinal sections, it is ascertained that the largest 

 vessels have their walls marked with both transverse and oblique lines 

 or bars, united together at the angles ; in some parts of the vessels, the 

 bars appear to cross each other, which gives them a reticulated appear- 

 ance. Brongniart assimilates these large vessels with some that are 

 common in Ferns and Lycopods. The smallest (between which and the 

 latter a regular gradation may be traced), possess true continuous spiral 

 fibres, coiled parallel to each other, and without any appearance of reti- 

 culation. Whether these are true spiral vessels, is a question over which 

 there appears to hang some doubt. From their shortness, and the pro- 

 bable existence of an appreciable membrane between the fibres, Brong- 

 niart says, that one might be disposed to conclude that they belonged 

 to that form of vascular tissue known under the name of ducts, or false 

 spiral vessels, and that they are intermediate to true spiral vessels, and 

 the scalariform tissue of Ferns and Lycopods. 



These vessels are found in the same position as the true spiral vessels 

 of dicotyledonous plants^ but in the latter they do not constitute bundles 

 so large, nor so well defined, as in Sigillaria elegans. In the Dicotyle- 

 dons, each bundle of spiral vessels corresponds directly with the inner 

 side of a ligneous bundle included between two medullary rays : often 

 the vessels are very small : when they are numerous, the smallest are 

 placed on the side adjoining the pith, or in the centre, and the largest 

 on that side next to the radiated or ligneous cylinder. In Sigillaria 

 elegans, however, each bundle of the medullary sheath embraces several 

 narrow ligneous bundles, intersected by a number of medullary rays, 

 and the arrangement of the vessels is the very reverse of that just stated. 



Thus, although these bundles have some resemblance, by their posi- 

 tion, to those of the medullary sheath of ordinary Dicotyledons, they 

 possess a character which distinguishes them from those of the latter 

 plants, and which as yet has not been met with in any vegetable form 

 now living. 



The ligneous cylinder, unlike the vascular circle, or medullary sheath, 

 is divided by narrow radiating spaces, possessing no remains of tissue, 

 but which radiating spaces, in all probability, originally constituted 

 the medullaiy rays. The vessels, or more properly speaking, tubular 

 cells composing this cylinder, are uniform as regards the markings on 

 their walls, and are rather variable in shape. A transverse section of a 

 single bundle intervening two medullary rays, often exhibits the orifices 

 of the vessels of a round, oval, or hexagonal form, clearly defined. 



