70 



SMITHSONIAN MISCELLANEOUS COLLECTIONS 



VOL. 70 



The opposition between the most eastern field, 0° to 9° west longi- 

 tude, and the most western fields, 20° to 29° and 30° to 39° west 

 longitude, is sharply indicated in the curves of figure 33, numbers 

 I, III and IV. Curve II for the middle field 10° to 19° west longi- 

 tude shows a transition form. 



The agreement between the yearly curves and the February and 

 March-April curves is distinctly indicated by taking the mean for 

 all four fields for February and March-April and also for the 

 years and comparing them as is shown in figure 34. The curve 

 of mean values for February and March-April combined, which is 

 drawn as a full line in the figure, shows particularly well the close 

 parallelism with the curve for the year (September-August). 



Figure 35 shows the variations of the yearly temperature (Sep- 

 tember-August) for Petersen's 1° fields. We note that the curves 



8 s mil 



Figure 34. Curves of the temperature mean for all four Danish fields 

 (compare figs. 31 to 33) for February, March-April (upper lines) and for 

 the whole year (lower lines). 



for the station No. i and w'estward to No. 7 show considerable 

 similarity each to each and form so to speak a certain type, which 

 however, gradually changes from the east toward the west. This 

 imparts to these curves an impression of trustworthiness. The 

 curves for the stations westward of station 8 have little or no 

 similarity each to each and this is very likely due in the greater 

 part to the accidental errors of the observational material. 



On the whole there is a certain similarity apparent between the 

 yearly curves for Petersen's stations i to 6 and the yearly curves 

 for the Danish fields in corresponding longitudes (see fig. 33). 

 The reader may compare figure 35, station i, with figure 33 I, 

 figure 35, stations 3 and 4, with figure 33 III, or figure 35, stations 

 5, 6, 7, with figure 33 IV. 



There prevails also a strongly marked similarity between the 

 yearly curves for Petersen's most easterly station, the curve for 

 February for the same station (see fig. 36, P St. I) and our curves 



