224 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I40 



greater differences between them than do the trees within a species 

 from the species average. The same appHes to XSC which has an 

 average departure of flush agreements within the separate trees of 

 19.6 percent in contrast with a species average of 75. 



The general habitat relationships of the trees of TTC, TTP, and 

 XSC are so identical that the differences among branches and trees 

 must be ascribed to variations in local, or microsite, factors or, in the 

 case of species, to specific differences. All the trees grew on the flat 

 upland surface of the High Plains. Those of XSC were located 5 

 miles from TTC and TTP. The XSC trees were not over 200 yards 

 apart whereas TTC and TTP on the College campus were within 600 

 yards of each other. The dominant summer rainfall comes in large 

 part as intense local showers that vary greatly in amount within a frac- 

 tion of a mile. All the evidence favors multiplicity, such as described 

 in this report, as a dominant characteristic of a lower forest-border 

 region where intense dry spells alternate with heavy rains during the 

 warm season. 



Steep slopes of other regions would tend to eliminate or mitigate 

 certain types of rainfall and thus perhaps reduce somewhat the in- 

 cidence of multiplicity. Differences in soil texture and differences in 

 rainfall regime other than to produce a constantly high soil-moisture 

 percentage would also modify the incidence of multiplicity either by 

 increase or decrease. As a matter of fact, the Lubbock area is not 

 alone in favoring multiplicity. 



Fluctuations of soil moisture may stand high in the list of factors 

 responsible for multiplicity but perhaps we must seek further for the 

 reasons responsible for the distribution of partial growth layers. Nu- 

 trients, especially as triggered by water supply, may be insufficient to 

 produce a growth layer as an entire sheath during a single intra-annual 

 growth flush. They may be sufficient only for a flush producing a par- 

 tial growth layer in certain branches or producing patches here and 

 there on certain branches. Thus, perhaps, there is an element of 

 "chance" as to which branch, or where on a single branch, a flush 

 produces a partial growth layer. 



The possible role of growth-promoting and growth-inhibiting sub- 

 stances must not be overlooked in regard to the localization of partial 

 growth layers within a branch and to the variation of multiplicity 

 among the branches of a tree. 



Uncertainty will exist until we know exactly why a cambium is 

 active at one place and not at another, and at one time and not at 

 another. 



