62 



CLIMATIC CYCLES AND TREE GROWTH 



cal methods of producing it. His own method interests us not merely be- 

 cause it produces an analyzing pattern or periodogram that measures con- 

 tinuous periods with precision but also because his "fine structure" displays 

 some basic elements of the cyclogram. 



Our automatic cyclogram discussed in the previous chapter is a pattern 

 using space distribution of the cycles in a common rectangular plot, derived 

 all at once from the extended data; Stumpff arrives at a pattern of somewhat 

 similar qualities by a timing device, that is, his pattern is not produced all 

 at once but is constructed photographically point by point. Stumpff 's 

 inspiration (like the author's) was Schuster's periodogram and a desire to 

 produce it automatically. We each developed a pattern of cyclogram type 

 and used it in producing the periodogram. The writer began using cyclo- 



Fig. 28— Plan of Stumpff' s periodograph. After Stumpff, slightly modified. 



gram results from the start (1913-14) and essentially abandoned the periodo- 

 gram after producing actual photographs in 1918. 



The fundamentals of Stumpff 's apparatus are shown in a slightly modified 

 copy of his drawing in figure 28. As I have sketched it here, R is an opening 

 through which passes a bundle of parallel rays that are converged on the 

 cylindrical lens C by a condensing lens L. K is the curve cut out of opaque 

 paper, and G is a grating made up of equidistant openings; each one in width 

 is 50 per cent of the constant distance between their centers. W is a cylinder 

 with a longitudinal slit ; within the cylinder a drum carrying a photographic 

 film slowly rotates. Removing K for the moment and holding G stationary, 

 each opening in G will cast an illuminated image upon W, the image being 

 elongated perpendicular to the slit. If we should set G in motion in the 

 direction of the arrow, these images (for which the name "sweep" was long 

 ago suggested) will move to the left along the slit. Now holding G station- 

 ary and replacing K and starting the rotation in W, the several openings in 

 G will trace longitudinal lines on the rotating film, lines that are perpendicular 



