NO. 10 SOLAR RADIATION AND WEATHER STUDIES ABBOT 73 



The varves are supposed to have been formed as follows : During 

 Pleistocene glaciation considerable melting of the surface of the ice, 

 as well as copious rainfall, took place during the summer of each year. 

 This produced glacial torrents which scoured the sides of the glacial 

 valleys and carried down sediment. Settling occurred in the quiet 

 lakes which at the foot of the glacier intercepted the torrential flow. 

 In such settling the coarser particles reached bottom first, and the 

 finer particles were superposed thereon. The settling took place mainly 

 in the colder months after the melting had greatly diminished and 

 snow rather than rain fell, so that the turbulent streams nearly ceased. 

 In this way each year a layer of sediment was deposited, coarser at 

 bottom, finer above, and layer after layer formed as the years succeeded 

 each other. 



Many thousand years have since passed. Many variations of pres- 

 sure, of hardness, of exposure, and of still other factors must be 

 supposed to have affected the thickness of varves, besides the warmth 

 and the rainfall, of which we are now to invoke them as the witnesses. 

 Hence we can not hope to find the 23-year cycle very sharply defined 

 in varve thicknesses. But it may be that by taking the mean values 

 over intervals of 1 15 years, covering five cycles each, as was done with 

 the tree-ring measurements, interesting results will appear. 



With this anticipation I read off from Reeds' plots the thickness 

 of varves for a continuous interval of 575 years, and arranged the 

 values in five tables of 23 columns and 5 lines each. In figure 31, I 

 give the results of that investigation. It seems to show that in Pleisto- 

 cene time, as now, a 23-year cycle in temperature and rainfall resulted 

 from the summations of the effects of periodic variations of the sun. 

 Eight crests which appear in the general mean seem to be present 

 almost without exception in very nearly the same phase in the five 

 constituent curves. The range of values plotted in the general mean 

 curve, F, is from 1.44 to 2.00. a range of 40 percent. The range in 

 curve A is from 1.02 to 2.22, a range of 120 percent. 



24. A Test of the 23-YEAR Cycle in Eocene Varves and 

 Tree Rings 



Dr. Wilmot H. Bradley, United States Geological Survey, was 

 so good as to furnish me with several sets of measures of varves 

 and tree rings relating to Eocene times. These data included a con- 

 tinuous series of varves from the Green River formation, Parachute 

 Creek, Colo. They appear to have been formed by the annual ex- 

 pansion and drying up of a lake bed. These varves each presented 



