944 



SCIENCE 



[N. S. Vol. XXX. No. 783 



struetive insight into the conditions of the 

 atmosphere well back toward its primitive 

 state. They challenge at once the view 

 that in those early ages the earth was 

 swaddled by a dense vaporous atmosphere 

 from pole to pole ; for under such a vapor- 

 ous mantle a great desert tract in India 

 would be scarcely credible. 



If we come forward in time two periods, 

 to the deposits of the Silurian stage, we 

 find that underlying the basin of the St. 

 Lawrence in New York and westward there 

 stretch great sheets of salt and gypsum, 

 many thousand square miles in extent. 

 These beds are accompanied by complete 

 barrenness of life in some parts, by pau- 

 perization of life in other parts, by selec- 

 tions of life according to tolerance of sa- 

 linity in still other parts, and by har- 

 monious physical characters, all of which 

 combine to add strength to the interpreta- 

 tion. All these imply a degree of aridity 

 approaching desert conditions in what is 

 now the well-watered region of our Great 

 Lakes. These signal facts join those of the 

 Salt Range of India of earlier date in chal- 

 lenging the former conception of a uni- 

 versal envelope of vapor and cloud in all 

 those early times. 



In the next period there are formations 

 that have been interpreted as implying 

 desert conditions, but perhaps on less firm 

 grounds, and we pass on to certain stages 

 in the sub-Cai'boniferous period next fol- 

 lowing, wherein beds of salt and gypsum 

 are found in Montana, Michigan, Nova 

 Scotia and Australia, which imply like 

 climatic conditions. If we pass on to the 

 Permian and Triassic periods, near the 

 middle of the geologic series, beds of salt 

 and gypsum are phenomenally prevalent 

 on both the eastern and western continents, 

 reaching through surprising ranges of lati- 

 tude. The relative paucity, as well as the 

 peculiar characteristics of the life of those 



times, seems equally to imply vicissitudes 

 of climate in which scant atmospheric 

 moisture was a dominant feature. There 

 seems no tenable way to interpret these 

 remarkable facts of the middle periods 

 except by assuming an even greater preva- 

 lence of aridity than obtains at the present 

 time. So, at times in the later periods, but 

 at times only, the stratigraphic record im- 

 plies atmospheres as arid as that of to-day, 

 not everywhere, indeed, but in notable 

 areas and in certain horizons. 



These and other significant facts of con- 

 sonant import form one group of phe- 

 nomena. 



If, on the other hand, the record be 

 searched for facts of opposite import, they 

 will come easily to hand. Starting near 

 the beginning of the record, it is even more 

 easy to find stages abounding in evidences 

 of prevailing humidity, of great uniform- 

 ity of climate and of most congenial life- 

 conditions reaching through wide ranges 

 of latitude. If we rested on this selection 

 alone, the old view would be abundantly 

 sustained ; but the strata bearing evidences 

 of aridity lie between these. Combining 

 the two sets of facts, the conception seems 

 to force itself upon us that from the very 

 earliest stages of the distinct life-record 

 onward, there have been times and places 

 of pronounced aridity much as now, or 

 even more intense, while at other times, 

 intervening between these, more humid 

 and uniform conditions prevailed. 



This conception grows in streng-th as we 

 turn from atmospheric states to prevailing 

 temperatures. The body of scientific men 

 have rarely been more reluctant to accept 

 any interpi'etation of geologic phenomena 

 than that of recent general glaciation on 

 the lowlands of Europe and America in 

 mid-latitudes when that view was first ad- 

 vanced by Louis Agassiz. With the con- 

 ception of former pervasive warmth then 



