154 



THE CLIMATIC FACTOR AS ILLUSTRATED IN ARID AMERICA. 



trunk which maintains its thickness well up to the top, so that for long distances the 

 diameter varies only at a very slow rate. At the base, however, as soon as the trees become 

 of any considerable age, the trunks flare and send out buttresses. The woodsmen who 

 fell the trees naturally prefer to do their cutting at a point above the flaring portion, and 

 thus save themselves the work of cutting through the extra thickness. In the case of 

 young trees, not over 1,000 years of age, this is feasible, although it necessitates the building 

 of a platform at a height of 6 feet or so above the ground. In the case of older trees the 

 amount of flaring is so great that it is impracticable to avoid it unless the trees are cut at 

 a height of 15 or 20 feet. Wlien the lumbering of the sequoias was first begun, this was 

 sometimes done, and in order to climb to the top of certain old stumps we were obliged to 

 throw ropes over them and clamber to the height of a second-story window. 



In later years, however, the practise has been to cut the trees low. Some of the best 

 and oldest stumps are cut not more than 3 or 4 feet from the ground, and 5 or 6 feet is 

 the ordinary height. In the flaring portion of a trunk the arrangement of the rings, as it 

 appears in a vertical section through the middle, is illustrated in figure 39, where it will 

 be seen that the central rings — those which 

 grew while the tree was young — are vertical 

 and the Une of cutting is at right angles to them ; 

 but the later rings have a distinct slope parallel 

 to the flare of the tree. Hence, when they are 

 cut horizontally by the saw of the woodsman, as 

 indicated by the solid Une AB, they are not tran- 

 sected at right angles. Thus their apparent 

 width, as seen on the surface of the stump, is 

 greater than their real width, as shown in the 

 dotted Hnes. In a few special cases we were 

 able to make allowances for this and to measure 

 the rings at right angles instead of horizontally ; 

 but in the majority of cases this was utterly out 

 of the question. It would have involved so 

 much cutting, so much extra time in measuring, and so much danger of making mistakes 

 in counting and measuring, that not only would the number of trees that we could measure 

 have been too small to give reliable results, but the character of the measurements would 

 have been more open to question than is now the case. The only course seemed to be to 

 measure horizontally and then to make corrections in the final curve. 



The degree to which this flare at the base of the trunk is effective may be judged some- 

 what from table 6. This table is compiled from sections of wood cut from relatively young 

 trees, for the purpose of making measurements of the growth during the last 50 or 100 

 years, by years instead of by decades. Trees having a diameter of not over 6 or 7 feet 

 near the base and not over 600 or 1,000 years old were selected; yet even these relatively 

 young trees flare considerably. Out of 85 specimens which happen to be at hand at the 

 time of writing, I have selected the 32 which show this effect most clearly. A glance at 

 the table shows that the horizontal distance along which measurements are made exceeds 

 the actual distance at right angles to the rings by amounts varying from 3 per cent to 16 

 per cent, the average being 7.6 per cent. Out of the 85 trees, 32 others, while flaring 

 noticeably, show a less degree of flaring— that is, 2 to 5 per cent— while 21 show the effect 

 of flaring to the extent of 2 per cent or less, and may be considered practically straight. If 

 the curve of growth of the 32 young trees whose flare has been calculated were plotted 

 without further correction, the part belonging to the present time would be nearly 8 per 

 cent too high in comparison with the earlier parts. With old trees this effect is exaggerated. 

 From this in part comes the rise in the latter part of the soUd line in figure 38. 



Fig. 39.— Effect of Flaring Buttresses on the 

 Measurementof Rings of Growth. 



