238 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 70 



ished. Hence we should expect that the temperature amplitude 

 would have the tendency to diminish at times of minimum mean 

 temperature ; for the heat which is communicated to tke surface 

 of the earth by the sun rays would vary more than the outgoing 

 radiation, although this also would be diminished by the formation 

 of clouds and by dust in the atmosphere. But any marked diminu- 

 tion in the daily or yearly amplitude of the temperatures at minimum 

 of mean air temperature is in general not to be found in our curves 

 or at least not to be found in that degree which would be expected. 



PROOF OF THE FAILURE OF BLANFORD's HYPOTHESIS SHOWN BY 

 OBSERVATIONS IN THE INDIAN OCEAN 



We have already remarked that the direct observations do not 

 support Blanford's hypothesis, namely, that in consequence of great 

 solar radiation at sun spot maximum the surface of the ocean is 

 warmer than at sun spot minimum, and therefore increased evapora- 

 tion produces greater cloudiness and more precipitation over the 

 land which finally produces lower temperatures. We will now take 

 up this point more in detail. 



According to the observations which the Dutch have published for 

 two 10° squares in the Indian Ocean between o° and io° north 

 latitude and between 70° and 90° east longitude, we have drawn the 

 curves in figure 90 for the anomalies of surface temperature (WT) 

 for the two fields combined (curves III and VIII) also for the 

 curves for the air temperature (T) fo'r the same two ten degree 

 squares (curves IV and IX), also the curves for the wind velocity 

 (W), expressed according to Beaufort's scale without regard to 

 the direction. These are curves VI and XI with scales inverted. 

 Finally we give curves for the cloudiness (N), the curves VII and 



XII also with inverted scales. On the same figure at the top we 

 have given in curves I and II the temperature (T) and the air 

 pressure (P) in Mauritius and at the bottom of figure the curves 



XIII to XV for the air temperature (T), the air pressure (P), and 

 wind velocity (W) for Batavia. 



We see from this figure that the variations in the surface tem- 

 perature and the air temperature in these parts of the Indian Ocean 

 follow one another closely and show also a great agreement with 

 the variations of the air temperature of Mauritius and Batavia. 

 There are a few exceptions, as for instance, that the maximum in the 

 air temperature of Mauritius in the year 1908 does not appear in 

 the curves for the two 10° squares in the Indian Ocean, nor does it 



