NO, I GROWTH LAYERS IN TREE BRANCHES CLOCK ET AL. 73 



Suppose we draw radial lines across 1939 of XSC i-i-b (text fig. 

 14), one passing through the left-hand letter C, another through the 

 right-hand letter, C, and a third through the letter Di. Then let us 

 translate the structural features appearing along these lines into cam- 

 bial activity. It is at once clear that the time of activity, the intervals 

 of activity, and the rate of activity do not coincide either on the three 

 radials taken together or on any two of them. After tracing the ac- 

 tivity along various radials, we can more readily obtain a picture of 

 general activity for 1939 over an area rather than on a cross section. 

 Specimens such as XSC i-i-b exhibit convincing evidence that cam- 

 bial activity is not a simple process which begins necessarily with a 

 burst of cell division in the spring, runs a regular course, declines, 

 and finally ceases, not to be disturbed again until the following spring. 



The section XSC i-i-b (text figs. 15, 16) shows the increment for 

 1940 to be a system of compound and overlapping lenses, a half-lens, 

 and two complete, entire growth layers. Actually two systems may be 

 identified : that including growth layers i to 5 and that including 

 growth layer 6 (text fig. 16). Growth layers i, ii, and 2 entail all the 

 problems already discussed in reference to half-lenses. Otherwise, 

 lens 4 overlaps lens 3. Lenses 4 and 5 are a true compound lens 

 wherein the outer, 5, covers the shorter arc (text fig. i). It is in- 

 teresting to note that, after the cambium indulged in activity of 

 highly variable rate and location, i to 5, activity began everywhere on 

 the cross section at about the same time, ran a uniform course ap- 

 parently, and ceased everywhere at about the same time. 



In connection with all the problems of variable cambial activity, the 

 information from the Director of the Agricultural Experiment Sub- 

 station is to be recalled : the trees used in our work were not irrigated 

 regularly after about 1936. 



One more example is given from section XSC 1-2-b (text figs. 17, 

 18), showing increments for 1938 and a portion of 1939. The frosts 

 for each year are shown at the start of each annual increment. Growth 

 layer i (text fig. 18) is complete and entire. In contrast, growth 

 layers 2 and 3 comprise a compound lens with the smaller lens on the 

 outside (text fig. i). At the midregion of the arc covered by growth 

 layer 2, its outer margin is slightly indefinite. Here cambial activity 

 had slowed down but not wholly ceased. In spite of this small region 

 of indefiniteness, the fact that lens 3 covers less area than lens 2 in- 

 dicates that lens 3 was made by a period of cambial activity distinct 

 in a very real sense from that which made lens 2. The portion of the 

 increment for 1939 shown in the drawing includes growth layers num- 

 bered 4-41-42, 5-51, and 6. Growth layer 6 is a simple lens. It is the 



