356 THE AMERICAN NATURALIST. [Vol. XXXVIII. 



to a substitution of functional activities between these structures 

 and the degenerate parenchyma cells. We may therefore con- 

 clude that extreme variation in the character of the pit is an 

 expression of a higher type of development, and that from this 

 standpoint, such structures have a definite value in solving ques- 

 tions of descent. 



The terminal walls of the ray cells present three variants with 

 respect to secondary growth. All the more primitive Cordai- 

 tales and Coniferales are characterized by thin walls. Cupressus 

 and Juniperus are chiefly distinguished by their thin walls which 

 are also locally thickened, a feature which has been shown to be 

 due to incipient secondary growth. But such alterations are 

 already foreshadowed in Libocedrus where the local thickening 

 of the wall is of a sporadic nature. In Abies magnifica and 

 A. grandis there is a partial recurrence of thin and locally thick- 

 ened walls, which is pretty fully expressed in A. concolor. A 

 similar recurrence is met with in Pseiidotsuga macrocarpa, in 

 Picea polita and in Piniis parrayana, and it is also complete in 

 thirteen of the most highly developed species of Pinus, where 

 the walls have suffered extreme degeneration. Within the 

 limits of Picea (i) and the soft pines (5), there are six instances 

 in all, of sporadic and partial survival of the thin and locally 

 thickened wall. The first tendency to thick and strongly pitted 

 walls is manifested in five species of Juniperus, and such 

 development is fully expressed in what may be regarded as the 

 three most highly developed species. Thick walls are then 

 fully characteristic of Abies — with a partial reversion in 

 A. concolor, of Tsuga, Pseudotsuga douglasii, Picea, with the 

 exception of P. polita, five species of soft pines and three species 

 of hard pines. In P. tceda and P. palustris the walls are so 

 degenerate that their structure cannot be satisfactorily deter- 

 mined, but they are presumably thin- walled. 



From these facts it is manifest that the progressive thickening 

 of the terminal walls accords with the general course of develop- 

 ment, and once more making use of the principles already 

 applied to the pits on the lateral walls, we are brought to the 

 natural conclusion that (i) an increase in the thickness of the 

 walls is evidence of a higher type of organization, and (2) that 



