122 



the Batavia Observatory, especially by Observator J. H. Kats, to 

 all of wiiom we lierevvftli desire to express our thanks 



In order to gather further knowledge of the air currents in the 

 lower layers, we collected and worked out particulars for each 

 100 m., up to J 500 m.; upwards of 1500 m. for each 500 m. 



The following is a synopsis of the number of cases on which 

 finally the normal-equations are based. In order to somewhat neutralize 

 the diminution of those numbers in the higher levels, groups were 

 formed for the deduction of the coefticients, resulting in the follow- 

 ing figures : 



This compensation was also necessary to meet the growing uncer- 

 tainty of the observation with increasing height and the increase 

 of the (non-periodical) velocity of the wind. 



However, the nuniber of cases of such a conibined group may 

 not be regarded as being equivalent to a non-combined one of the 

 same number, seeing that the non-periodical velocities of the wind — 

 here appearing as accidental errors — are not independent of each 

 other for successive heights. 



It is, however, impossible to ascertain this difference quantitatively, 

 although it should be remarked that it is principally dependent on 

 two circumstances. 



In the first place on the variability of the wind which the 

 balloon encounters when ascending. As a rule this variability will be 

 somewhat less in the higher layers than in the lower ones. 



Secondly on the manner of combining the ascensions, i. e. whether* 

 the differences of the wind have been deduced from (he same couple 

 of ascensions or from various ones. E. g. a balloon sent up at 3 a.m. 

 is compared up to 6 km. with one sent up at a.m., but higher 

 up with one sent up the day before at 6 p. m. In this case the 

 differences for e.g. 4.5 km. and 5 km. will be more dependent on 

 each other than those for 5.5 km. and 6 km. The letter case will 

 be more frequent in the upper layers. 



