CYCLOGRAM ANALYSIS 47 



a large movable rack in which several cycleplots may be placed one above 

 the other. By aid of the rack, each plot in turn can be brought directly in 

 front of the illuminated ground-glass and analyzed. The plots are arranged 

 so that their dates coincide ; hence in a second's time one can change from the 

 inspection of one record to that of another. This makes the comparison of 

 cycle characters in different records very rapid and accurate. For instance, 

 the individual curves of trees in a group can thus be placed one above the 

 other and the group cycles derived by the rapid comparison of the different 

 records. Comparisons are so quickly made that Mr. Schulman has often 

 called the process "simultaneous analysis" because the time interval between 

 the views of successive patterns is almost negligible. 1 The rack was first con- 

 structed about 1921 in an attempt to carry dating by cycles through Arizona 

 trees and into California sequoias. In a considerable number of tests we 

 find that results are closely alike, whether by analysis of a group curve or 

 the separate analysis of the individuals composing the group. A comparison 

 between the results of using these two methods is shown in figure 35. 



OPERATION AND CHECKING 



In performing analyses with the cyclograph, the outward movement of 

 the mirror causes rows of maxima to rotate in a clockwise direction. When 

 any row reaches a horizontal position indicated by a stationary thread in 

 the focal plane the reading is taken (see Plate 15B). It is standard practise 

 to note this reading of cycle length while moving the mirror outward and 

 to repeat the reading on moving it inward. The different approach some- 

 times causes slight change in estimate of the true setting. Record is kept 

 of both readings and the mean taken. 



According to the importance of the case, further checking is made. The 

 cutting line is put on an inverted curve and cycles are accepted as valid that 

 are common to the two tests. It is, of course, most desirable to have someone 

 else try the analysis, but for years until very recently that was impossible. 

 In 1926 the first wholesale analyses were made of the 42 western groups and 

 certain general characters of apparent importance were noted and needed 

 checking. Before publication of Volume II, Climatic Cycles and Tree Growth, 

 in which the results had been tentatively placed, resort was made to checking 

 by replotting at unknown scale and making the analyses myself. This was 

 done and results obtained were incorporated as a foot note on page 124. This 

 check at unknown scale is given in Reports of Conferences on Cycles, 

 Carnegie Institution of Washington, 1929. 



Accuracy and Limitations — During the twenty years of use of the cyclo- 

 graph, checking has been done as described, but the need of determining 

 fundamental errors, limitations and capacities of the instrument, was fully 

 realized. Formal tests along these lines only recently became possible. Mr. 

 Edmund Schulman several years ago became skilled in the use of this instru- 



1 We are also using the word "concurrent" but expect to make the word "simulta- 

 neous" literally applicable later on. 



