164 AXXALS NEW YORK ACADEMY OF SCIENCES 



8°. 14, 8°. 17 and 8 ".02 C. Curve C indicates a long range variation 

 completely independent of the temperatnre variation, which has a maxi- 

 inuin in 1902 and a minimnni in 1910, as can he seen on cun^es A and 

 B. The data which I collected some years ago from the observations of 

 Eussian stations^^ give the result that these long-range variations of the 

 values of the mean diurnal oscillations of temperature differ from one 

 region to another and that these variations correspond to changes of 

 cloudiness. An increase of cloudiness diminishes the average daily range 

 of temperature. Considering now the trend of the curve we notice a 

 more or less progTessive increase of the values from 1901 to 1904. This 

 is just the contrary of what would be expected. At Fort-de-France the 

 daily range of temperature has therefore not been greatly affected by the 

 dust veils produced by the eruptions of Mt. Pele, Sta. Maria and Mt. 

 Colima. 



The mean temperatures have also been affected but very slightly. The 

 jileionian crest of 1902-1903, as indicated on the curve of consecutive 

 means, has been depressed a little, but certainly not more than 0.15 °C. 

 or 0.2 °F. It is difficult to judge how much the mean temperatures of 

 the individual months have been affected. Curve .1 shows a decrease of 

 the departures for the months of June, July and August. 1902, following 

 the great Mt. Pele eruption : also a more pronounced decrease for Novem- 

 ber and December, possibly due to the Sta. ^laria eruption, and finally 

 the low April departure may have been caused by the Colima eruptions. 

 However the departures of the months of May, 1902, to the end of 1903 

 are all above the average and if the slight deflections observed during the 

 period of great volcanic eruptions must really be attributed to dust veils, 

 it may be presumed that the means of some months have been affected 

 more than those of other months but none sufficiently to mask the plei- 

 onian character of the departures. Moreover, the effect of the dust veil 

 ceased long before the complete development of the antipleionian de- 

 pression of 1904-1905. This antipleion cannot, therefore, be considered 

 as being a consequence of the formation of the volcanic dust veil. 



Temperatures at Para, Catexxe axd the West Ixdies 



Although the different examples I have given may be considered as a 

 sufficient proof of the fact that the pleionian variations are absolutely 

 distinct in their origin from the possible temperature variations due to 

 the presence of volcanic dust, I will examine a few more consecutive tem- 



^^Henryk Arctowski : "Xotice sur les variations de longue dur^e des amplitudes 

 moyennes de la marche diurne de la temperature en Russie." Bull. Soc. bel.^e d'Astrono- 

 mie. Bruxelles, 1908. 



