320 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 3 



has been arranged on the graph spacing used in Figures 2 to 13. 

 This arrangement permits the entire smoothed curve of 736 varves 

 to be placed on one sheet. The annual varves do not appear. It 

 also affords an opportunity to see at a glance the more prominent 

 major fluctuations and periodicities in the melting of the ice. Regu- 

 lar cycles are not apparent. The first prominent warm period was 

 for the years 52-68. This was followed by a comparatively cold 

 period to the 112th year, but it was not as cold as during the next 

 192 years, terminating with varve 300. For eight years, 300-308, 

 the melting of the ice almost reached the normal between cold and 

 warm temperatures. From 308 to 332 the curve moves gradually 

 downward, indicating a 24-year period when the successive summers 

 were progressively colder. Then, during the next 204 years, there 

 follows a series of short cycles of warm and cold periods of vary- 

 ing length, which were for the most part below normal, terminating 

 with 528 years. Then, curiously enough, the last 208 years, 528 to 

 736, were cyclic in form, but predominantly warm — that is, above 

 the normal or mean line. When an opportunity has arisen to check 

 varves 52 to 68 and 700 to 716 with varves from other localities, 

 then we shall know whether these abnormal fluctuations were caused 

 by excessively warm periods or were accentuated by local causes. 



PERIODICITIES IN VARVE CLAY GRAPHS AND TELECONNECTIONS 



Periodicities in varve graphs have been noted by W. Koppen 

 (1928), C. E. P. Brooks (1928), and E. Antevs (1929). Koppen 

 recognized periods of 2, 3, 4, 5, 6, 7, and 8 years in North American 

 varves, with the period of 2, 3, 4, and 5 years being frequent. Brooks 

 has analyzed the teleconnections of De Geer between Sweden, North 

 America, Argentina, and India. In Argentina he notes cycles of 

 5.1, 10.4, and 51 years in a curve secured by Doctor Caldenius at 

 Lago Corintos, which was described and figured by De Geer (1927). 

 The period of 10.4 years noted by -Brooks approaches the 11-year 

 sun-spot curve in length, but not in rhythm. In fact, Brooks com- 

 ments on the almost entire absence of an 11-year periodicity in the 

 varved clays. Antevs states that while stadial moraines indicate 

 the periodical retreat of the ice edge and are especially suited for 

 long-range periodic phenomena, the clays enable us to determine 

 the exact length and character of the periods. The varve graphs 

 are best suited for the study of short cycles. He regards the al- 

 most complete absence of the 11-year cycle in the curves studied 

 by Brooks as being perhaps the most important result so far obtained 

 from the analyses of varve curves. 



Brooks (1928), in discussing De Geer's (1926) teleconnections 

 between Zealand (Denmark) and the Little Ferry, N. J., clays of 

 the Hackensack Valley, which were described and diagramed by 



