t^te Bolton Balhvai/^ at Dixon Fold, near Manchester. 157 



upper surface, or that most influenced by light and heat. Of 

 a similar character are the swollen base and spreading roots 

 of these fossils, the roots diverging downwards at an angle of 

 29'' with the horizon. But the soft monocotyledons to which 

 they have generally been referred, have a very different eco- 

 nomy ; palms and arborescent ferns grov upwards only (not 

 laterally) and from within ; and instead of the massive forked 

 and spreading roots of ordinary forest trees, have usually a 

 dense assemblage of fibres like those of an onion or a hyacinth. 

 The delicate straight or curved striae ?een on good decorti- 

 cated specimens of Sigillaria, are so similar to those on the 

 alburnum of some modern trees as to render it probable that 

 the fossils had, like them, a separate bark, a character consi- 

 dered by vegetable physiologists, as proof of a woody struc- 

 ture. The scars also left by the disarticulation of the leaves 

 are indicative of a dicotyledonous, if not of a wooded structure. 

 Analogy, therefore, is in favour of these fossils having been 

 solid timber trees. 



On the bank of a coal-pit in the same neighbourhood, I 

 found a portion of the trunk of another and similar fossil tree, 

 principally filled with shale, but having a portion of the in- 

 terior adhering to the side, which, on being sliced and polish- 

 ed, exhibited woody structure. My friend, Mr Robert Brown, 

 kindly undertook to direct proper sections to be prepared, by 

 means of which that illustrious botanist ascertained that in 

 the transverse section there was that uniformity of vascular- 

 ity which is evidence of the coniferous structure. In the 

 longitudinal section taken parallel to the medullary rays, the 

 existence of these rays was ascertained, so that the specunens 

 exhibit proof of dicotyledonous structure, and considerable 

 probability of that structure being coniferous. But the more 

 important evidence of discs in the section taken panillel to 

 the medullary rays was wanting, the vessels having apparently 

 undergone some alteration. This specimen, therefore, inde- 

 pendent of the question whether its exterior marldngs cor- 

 respond with those of the true Sigillaria or not, establishes 

 the important fact that some of those treeS which are believed 



