No. 449-] NORTH AMERICAN CONIFERALES. 



ential value than is at present apparent. A simple illustration 

 will serve to afford an idea of the rather limited specific value of 

 this character. In Taxus cuspidata the rays are numerous, while 

 in the two remaining species they are relatively few. The .same 

 feature applies to the differentiation of Torrcya mtcifcya from the 

 other species of that genus. In Pinns clausa, P. 

 scrotifia, P. murrayana, etc., the same rule applies, 

 but in all such cases it cannot be accepted as final. 



The height of the ray is subject to such varia- 

 tions, even within the same species, that it cannot 

 be defined with sufficient accuracy to admit of its 

 application to classification in more than a very 

 general sense. It is true that the rays of Ginkgo 

 are always low, while those of Taxus and Torreya 

 are often high. In Juniperus they are commonly 

 low, while in Pinus they range from low to very 

 high. Such variations do not possess sufficient 

 constancy to admit of either generic or specific 

 application in the strict sense, though they not 

 infrequently serve a useful purpose as controlling 

 factors, and they are therefore incorporated in all 

 the diagnoses. Variations in breadth have a much 

 more definite value, since the element of constancy 

 is well defined. The genus Thuya (Fig. 31) may W 

 almost invariably be differentiated by this feature. H 

 In Cupressus, C. thyoides may be distinguished by 

 a similar feature, while C. arizonica and C. govciii- ru.. ii -Thuya 

 ana are equally well indicated by the great breadth gl'ItiirsectL"n 

 of the rays. The same rule applies also to Juni- hlVhTt^'icIi 

 perus, Sequoia (Fig. 25), Pinus, and other genera, ly narrow and 

 whence it appears that this feature is of specific J'X^''^"'' 

 value. It is always associated with and dependent 

 upon the form of the component cells (tangential) which afford a 

 means of distinguishing genera and species with much directness. 

 The narrowly oblong cells of Thuya (Fig. 31) serve to separate 

 this genus without difficulty, since a similar feature occurs but 

 rarely elsewhere, and then in such association as to make the 

 differentiation clear. In Juniperus the genus is separable into 



