OF CENTRAL CANADA PART IV. 



217 



ship of Bosanquet on Lake Huron. Other examples occur in the 

 Devonian rocks of Gaspe. At the latter locality, some impressions 

 of radiating leaves (annularia laxa, Dawson) belonging probably to 

 a related type of plant, have also been found. The narrow radiating 

 leaves (often attached to furrowed stems) known as Aster 'ophyllitesi 

 and which are so abundant in many Carboniferous and in some 

 Devonian strata, have not yet been recognized within the area of 

 Ontario and Quebec. They are commonly regarded as calamite 

 leaves, but on very uncertain evidence. 



Ferns (Filices), although so abundant in the higher Devonian, 

 Carboniferous and other strata, mostly in the form of leaf or frond 

 impressions, have not as yet been discovered in a fossil condition 

 within the limits of the Provinces referred to in the present work. 



The Hydropteridce or Rhizocarps, sometimes known as water-ferns, 

 comprise merely a few fresh-water forms (Marsilea, Piluria, Salvinia, 

 &c.) without fossil representatives in our strata. 



The Lycopodiacece, of existing Nature comprise a small number of 

 inconspicuous forms belonging to both land and fresh- water types. 

 The former (lycopodium, selaginella) are small, moss-like cryptogams, 

 dichotomously branching, and with narrow, more or less clasping or 

 imbricating leaves. The aquatic types (Isoetes) are rush-like forms. 

 True lycopods occur in Devonian and higher strata; and some 

 apparently related forms from the Devonian rocks of Gaspe have 

 been referred by Sir William Dawson to this division. The principal 

 of these form the genus Psilophylon (Fig. 120), represented by frag- 

 mentary impressions of narrow, stem-like plants, circinate (as in 

 ordinary ferns) at their terminal points. The leaves are very small 

 and thorn-like. 



The Lepidodendracece, represented typically by the fossil genera, 

 Lepidodendron and Sigillaria, are entirely Palaeozoic. They consist, 

 in most cases, of casts or impressions of tree-stems, usually frag- 

 mentary, but found occasionally in lengths of more than thirty or 

 forty feet. The lepidodendracese, proper, are regarded as closely 

 allied to the lycopods, whilst the sigillaria? are thought by some 

 authorities to constitute a higher type of vegetation, more nearly 

 allied to the cycads. The two, however, are very closely alike,* and 



* In their dichotomons branching, their supposed leaves, their leaf-scars, woody structure, 

 roots, &c. See a comparative tabular view of these homologies by Professor Schimpfer, in 

 Zittel's Handbuch der Palaeontologie (1880), p.p. 209, 210. 



