88 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 70 



The average values for the great regions (see curves W and L 

 in fig. 48) show the variations of the air temperatures and the sur- 

 face temperatures in good agreement, although the variations of the 

 air temperatures are considerably greater that those of the ocean 

 surface. This is particularly the case in February in the middle 

 region of the ocean as shown distinctly in figure 49. In plates 42 

 to 45, we have given the curves for the temperature of the air (L) 

 and of the water (W) for the separate 10° longitude fields. We 

 find there particularly in February the same tendency to considerably 

 greater variations (with maxima and minima more strongly de- 

 veloped) in the air temperature than in the surface temperature, 

 yet there are many exceptions in the two most easterly as well as in 

 the two most westerly fields. 



There are, however, certain marked disagreements between the 

 curves for the air temperature and the curves for the surface tem- 

 perature of the ocean. This can be observed in the February curve 

 for the air temperature in the middle of our fields along the route 

 Channel-New York (see fig. 48) and also in the mean values for the 

 middle fields of figure 49, which show a secondary minimum in the 

 year 1907 which finds no place in the curves for the surface tempera- 

 tures. In some of the curves for the fields in the region Portugal to 

 the Azores we also find a tendency to a similar secondary minimum 

 in the curves of air temperatures (see fig. 52). Since it appears in so 

 many curves for different fields, particularly for the 10° fields 30° 

 to 50° west longitude in the route Channel-New York, we cannot 

 think that this depression is merely accidental, but rather that it 

 probably corresponds with a real condition. 



The March-April curve for the air temperature for the average 

 of the fields on the route Channel to New York (see fig. 48), or 

 for the four middle 10° longitude fields (fig. 49) show a remarkable 

 rise of temperature for 1904 in relation to 1903 and 1905. There 

 is nothing corresponding to this in the surface temperatures. Since 

 this noteworthy rise of air temperature in the year 1904 is appearing 

 in all the curves for the air temperature in the 10° fields between 

 20° and 60° west longitude and most strongly so in the midmost 

 of these fields, that is, the field between 40° and 50° west longitude 

 (see pi. 42), we have to do in this case certainly with the real rela- 

 tion of things and not with mere accidental errors in the observa- 

 tions. There are still other traces of disagreement between the air 

 temperature and the surface temperature, of which we shall speak 

 later. 



