NO. I GROWTH LAYERS IN TREE BRANCHES CLOCK ET AL. 255 



on one section changes within several centimeters to an arc of frost 

 injury, or a series of separated arcs, whose circuit is completed by 

 densewood. Farther along the branch, frost effects disappear entirely 

 and leave a continuous band of densewood. Such transitions have 

 been noted particularly in the apple, the Arizona cypress, and the pine. 

 One of the best examples was noted in XSC 9-5 where 9-5-a, 52.5 

 cm. from the tip, contains a complete circle of frost in 1938, with 

 which densewood cells are associated. In 9-S-b, at 48.5 cm. from the 

 tip, no change has occurred. In 9-5-c, at 42.5 cm. from the tip, the 

 frost effects are much subdued and the band of densewood is com- 

 plete and entire. In 9-5-d, at 33.5 cm. from the tip, direct frost injury 

 and recovery are absent but are represented by diffuse densewood. 



Resin canals^ compression wood, and parenchyma cells are asso- 

 ciated with, or substituted for, frost injuries in the same manner as 

 the densewood cells. The first two occur rarely but the parenchyma 

 cells are as common, although not as continuous, as densewood. Com- 

 monly, an arc of frost injury is completed around the circuit both by 

 densewood and by parenchyma cells. Thus these last, like densewood, 

 appear to indicate an interruption in the normal course of growth not 

 long after the start of growth in the spring. Many annual increments 

 not only have parenchyma cells connected with frost injuries but also 

 have circles of such cells farther out in the xylem (at times associated 

 with densewood) which was formed much later in the season. Per- 

 haps it is reasonable to suppose that these circles of parenchyma cells 

 are also caused by some interruption in the normal course of growth. 

 They have been studied especially in the Arizona cypress, the cotton- 

 wood, the Siberian elm, and the juniper. In a longitudinal direction, a 

 circle of parenchyma cells may represent the frost injury existing else- 

 where on the branch. Within a space of 20 cm., on XSC ii-i (pi. 30, 

 fig. i), a complete circle of frost injury in 1938 becomes very much 

 subdued and is replaced by densewood and parenchyma cells. In 

 branch XSC 11-3, the frost injury is entirely absent leaving only the 

 densewood and the parenchyma cells (pi. 31, fig. 2), all within a dis- 

 tance of 3 cm. The use of parenchyma cells as a dating medium in 

 place of the frost injury they are supposed to represent must be car- 

 ried out with great care, if at all. 



If frost strikes before the cambium becomes active around the en- 

 tire circuit of a branch, an interior lens may be formed. Frost- formed 

 interior lenses are fairly common and have been observed chiefly in 

 Arizona cypress (for instance, XSC 1-2, 1-5, 4-1, 13-3) and in pine 

 (TTP 21-3, 24-10, 24-14, XSP 2-2, 2-3). In XSC 1-2, a frost injury 

 is directly associated with the densewood of the lens. In XSC 1-5, the 



