34 EOYAL SOCIETY OF CANADA 



Pleistocene may have its structure perfectly preserved, another lignite 

 from a much more recent deposit may show but few of those structural 

 features upon which distinction of species may be supposed to rest. From 

 this it follows that, assuming such a classification to be perfectly elabo- 

 rated, its application to the determination of fossil plants must have very 

 definite limitations, dependent upon the state of preservation of the ma- 

 terial in hand. 



In accordance with these considerations, it was originally held that 

 anv such classification, to be most efficient for all purposes thus indicated, 

 must permit conclusive deductions to be drawn, if possible, from sections 

 of about one centimetre square, such as might be prepared in the ordin- 

 ary way for microscopic purposes, since this alone would meet the aver- 

 age requirements of material representing fossil plants. It is to be ob- 

 served, however, that such limitations at once impose difficulties which, 

 joined to those due to the fact that the wood alone furnishes the necessary 

 data, might render the classification of inferior value in actual practice. 

 The aim has, therefore, been to select if possible, those distinguishing 

 characters which may be found in the structure of the woody parts of the 

 stem as exposed in the usvial planes of section — transverse, radial and 

 tangential, — and to obtain conclusive proof as to their efficiency or ineffi- 

 ciency for the purpose stated. The results so far reached seem to justify 

 the conclusion that for genera th© characters are well defined. In this 

 connection we may cite the results already reached by Goi:»pert,' and the 

 fact that an extended examination of woods led Solereder'^ to the conclu- 

 sion that the characters to be met with are sufficiently constant to admit 

 of distinguishing families, tribes, genera, and even species. 



With these thoughts in mind, attention was directed in the first 

 instance, toward the accumulation of authentic material — a work of slow 

 progress, now extended over a period of sixteen years, and, with respect 

 to some of the angiosperms, not yet completed. Within recent date, 

 however, all the various species and varieties of the Taxaceœ and Coniferœ 

 north of the Mexican boundary have been brought together. This fact, 

 joined to the pre-eminent jwsition occupied by these plants with respect 

 to their economic importance and pahcontological relations, led to their 

 being regarded as subjects best suited to immediate investigation. The 

 present paper, therefore, which deals with the gymnosperms only, may 

 be regarded as the first of a series of similar treatises on the classification 

 of the North American woods as a whole. 



Incidentally to the ])resent woi'k, a number of foreign species have 

 been studied, but it has been thought advisable to defer thc^ir consider- 

 ation until opportunity offers for an exhaustive treatment of all exotic 

 species. 



1 Foss. Conif., Leiden, 1850. 

 ■Î Bot. Zeit., xliv., 1886, oOfi. 



