NO. I GROWTH LAYERS IN TREE BRANCHES — CLOCK ET AL. 89 



This is amply illustrated by the increments for 1939 and 1940 in TTP 

 21-6-3., for 1939 and 1940 in TTP 21-8-a, for 1940 in TTP 21-8-b, 

 and for 1940 in TTP 23-3-a, Under low power, 1939 and 1940 ap- 

 pear to be very weakly developed lenses (actually divided densewood) 

 of one or several concurrent members. High power shows that it is 

 only the lightwood which is lenticular, the densewood being continu- 

 ous around the circuit. The entire band of densewood may be re- 

 duced at places to a thickness of one cell only. In fact, continuous 

 tracing under high power indicates that the combined bands of dense- 

 wood for 1939 and 1940 at one spot in TTP 23-3-a are reduced to a 

 thickness of two narrow cells. More than this, in TTP 23-i-a, the 

 densewoods of 1941 and 1940 lie against that of 1939 (see pi. 19, fig. 

 I ) . No evidence has come to light throughout the complete range of 

 the work to suggest that the cambium failed to divide during a single 

 season. 



The above examples do not show the most extreme cases of in- 

 cipient cambial activity. A study of the rather wide densewoods, as 

 well as the compression zones, of XSC 1-5-b, reveals many detailed 

 alternations of narrow and wide cells. Do these show high sensitivity 

 to slight fluctuations of physiological activity or repressed indications 

 of the impact of rather intense fluctuations ? The very beginnings of 

 divided densewood are shown in the broad band of 1937 densewood 

 of TTP 20-6-a which is made up of successive alternations of radially 

 narrow and wide cells. Here again sensitivity appears to have been 

 high, giving a rapid alternation of conditions so slightly recorded in 

 the xylem. 



Some evidence exists for the presence of what could be called di- 

 vided lightwood (text fig. 29b). Halfway out in the 1941 increment 

 of Con T 2-9-a, a zone of wide cells encircles the section and may 

 record accelerated growth at the start of the second 1941 tip flush (see 

 1933 in pi. 2, fig. 2). On the long radius and confined to the compres- 

 sion wood, the increment for 1942 in Con T 2-6-a gives evidence of 

 three diameter flushes : one-third and two-thirds of the way out there 

 are rows of wider cells. These seem to be a rather delicate test for the 

 presence of factors periodically causing accelerated diameter growth, 

 more delicate at least than tip growth of which there were two flushes 

 in 1942. 



Divided densewood and divided lightwood, as transitional forms 

 and as indications of cambial activity, are perhaps not only the most 

 interesting but also the most suggestive forms of partial growth layers. 



The term interrupted densezvood is applied to that form of partial 

 growth layer wherein the densewood, in its normally progressive 



