46 CLIMATIC CYCLES AND TREE GROWTH 



A cutting line should be made as nearly straight as possible or it may con- 

 fuse the analysis. This requirement was learned by experience. Certain 

 analyses made before 1919 showed the Hellmann cycle, a two-crested 11-year 

 cycle, very plainly (vol. I, Plate 12f). Analyses made early in 1932 failed 

 to show it. A test on the early cycleplot revealed that a 19-year and even 

 a 14-year cycle had been partly removed by a curved cutting line. To 

 verify this fact the cutting line was transferred to a cycleplot, analyzed, and 

 these two cycles were found in it. So, to extend the test, 19- and 14-year 

 cycles were subtracted mathematically from the Flagstaff data and the Hell- 

 mann cycle was once more revealed. It is legitimate to remove cycles in 

 the cutting line, provided the cutting line is itself analyzed to see what has 

 been removed. In general analysis such usage is not advised, but in special 

 cases this seems an advantageous way of removing interfering cycles. 



A straight cutting line aids the evaluation of longer cycles by leaving 

 them unaltered. The question of the cutting line then becomes a part of 

 a larger practical policy dealing with the range of the instrument as to cycle 

 length. The range extends from 5.0 years at the minimum to 30 or 40 years 

 at the maximum. Beyond 40 years we replot at a compressed horizontal 

 scale. At 30 years the visible cyclogram produced in the analyzing instru- 

 ment is reduced in length to f of its size at 5 years. Details are lost and the 

 larger cycles become visible. Therefore we must make a cycleplot that is 

 usable at each of these extremes. The nearly straight cutting line gives us 

 the long cycles; then not to lose the short cycles, obvious minima not deep 

 enough to reach down to the cutting line are joined to it by narrow "tongues" 

 of opaque paper that are visible at short range but disappear at long range. 

 One has only to try this to see that it does not distort the result. A cutting 

 line with these details is shown in figure 23. 



On the possibility that the cutting line had jeopardized correct analysis, 

 many important curves were plotted again upside down and a new cutting 

 line put in, encroaching this time a little on the maxima, and the minima 

 were cut out and tested. This was not found to change the results and the 

 cutting line procedure is considered satisfactory. It is evident that in this 

 analytical process the centers of mass of the maxima become the important 

 points of the plot whose periodic character is tested. In some cases the curve 

 has been modified into a series of centers of mass and the resulting cyclogram 

 becomes very brilliant and easy to read. In such cases, of course, caution is 

 needed and results obtained in this way are taken as corroborations or sug- 

 gestions (see cyclograms in Plate 15B). 



The instrument as it stands covers considerable space; namely 6 or 8 feet 

 in width by about 45 feet in length. The mirror is suspended from a track 

 attached to the ceiling. The cycle scale is marked on the track after calibra- 

 tion by standard curves. Calibration readings are made in connection with 

 every analysis. It is difficult to show much of this in any photograph but 

 something of it appears in Plate 14 A and B. 



Moving Rack— Many attachments to the cyclograph have been tried, 

 including moving picture devices, but the one that has proved its worth is 



