254 THE AMERICAN NATURALIST. [Vol. XXXVI 11. 



angle is 70 degrees, but the spirals are always characterized by 

 lack of prominence, they are often widely distant, and the some- 

 what extended areas within which they are wholly wanting or 

 fragmentary, suggests a process of obliteration. In P. doiiglassi 

 the average angle is 82 degrees ; the spirals are characterized by 

 considerable prominence and they are also, on the whole, close. 

 ' ■ . genus these two factors obviously possess a well defined 



differential value with 



respect to the two specie 



Among the higher genera only two cases are known in which 

 spirals occur, but in each the character is of a very sporadic 

 nature. In Larix americana spirals are frequently found in the 

 summer wood, but they are so inconstant in their occurrence, 

 and they present such varying aspects, that the angle cannot be 

 determined. In Finns tteda, where the walls of the summer 

 tracheids are very thick, rudiments of spirals may sometimes be 

 seen. Here also it is manifestly impossible to determine the 

 angle. 



Viewing these five genera collectively, their spirals conform 

 fully, in their occurrence and relation to progressive develop- 

 ment, to the general principles already stated, and especially as 

 formulated by De Bary. They possess no differential value of 

 generic rank with respect to Pinus and Larix, but they do have 

 such value with respect to Torreya and Taxus on the one hand, 

 and Pseudotsuga on the other, the differentiation resting upon 

 their occurrence in the summer wood in the former, and their 

 exclusion from that region in the latter. Were any question to 

 arise in this connection, it could be authoritatively decided by 

 the definite association of resin passages and fusiform rays in 

 Pseudotsuga. 



It only remains for us to ascertain how far such structural 

 eatures may be employed as a basis upon which to determine 

 the general phylogeny of the genera. As between Torreya and 

 laxus there is very little upon which to base conclusions 

 respecting sequence in development, and it is apparent that both 

 oi these genera have attained to nearly the same level. Such 

 differences as do exist, h..... ...^ 1 ^^j^,^,,. 



ugh slightly, inferior 



position of Torreya as indicated by (i) 

 genus and (2) the generally more com- 



