40 ANNALS NEW YORK ACADEMY OF SCIENCES 



with depressions below the average and elevations above, the knowledge 

 of the extent of the areas covered by positive and negative departures is 

 evidently more necessary for the discussion than the figures for some 

 isolated stations, where the temperature conditions may or may not cor- 

 respond to the average conditions of the surrounding countries. 



The wofk (lone for a previous publicatio;i was the mapping of all the 

 teni])('rature data F could obtain for the years' 1891 to 1900. 



Considering the means of the decade, 1891-1900, as being qua.si- 

 normal values, I have formed for each year and each station the depar- 

 tures from these means. These annual departures have been inscribed 

 on maps, and equideparture lines have been drawn. The areas of positive 

 departures have been called thermopleions, the areas of negative depar- 

 tures, thermomeions or antipleions. The result of the discussion is that 

 the year 1900 was a year of predominant thermopleions, the year 1893, 

 on the contrary, a year of most predominant antipleions. Taking the 

 probable areas into consideration, as well as the probable excess and 

 deficiency of temperature, I found that the difference in temperature 

 between these two years must have been at least 0°.5 C. 



This was the main result of my memoir "L'enchainement des varia- 

 tions climatiques," - published in 1909. This practical demonstration 

 of the fact that the temperature of the earth's atmosphere does not remain 

 constant leaves a very important question open for discussion. 



The annual departure maps of successive years showed in many cases 

 some striking similarities in the mutual relationship of pleions and anti- 

 pleions. I presumed, therefore, that pleions might persist from year to 

 year and that they displaced themselves. In my further researches, I 

 found it necessary to simplify the reasoning by adopting a way of ex- 

 pressing graphically the change of a given annual moan into that of 

 the following year. To avoid the more or less regular annual variation, 

 we have to take yearly means, but it makes no difference how we count 

 the year, as long as we compare means of 12 consecutive months. By 

 making consecutive yearly means for the one-year periods beginning with 

 January, February, March and so on, and by comparing the curves ex- 

 pressing the succession of the figures, we can see how a negative departure 

 of a given year passes progressively to a positive departure of another 

 year. 



I have published such consecutive temperature curves for the entire 

 series of observations recorded in Batavia^ and "Mew York,* and portions 



2 Henhyk Arctowski : L'enchalnement des variations climatiques. Bruxelles, 1009. 

 » Op. cit., p. 32. 



* Hknryk Arctow-ski : On Some Climatic Changes Recorded in New York City. Am. 

 Geog. Soc. Bull., Vol. 45, p. 117. 



