( WK ) 



which bind the electrons to their equilibrium positions must there 

 fore also be of electromagnetic nature. Thus here the only force 

 to be considered, is electrostatic, we know from the theory of 

 potential that an electrostatic tield permits of a point-charge moving 

 about a stable position of rest only inside of <i charge of opposite 

 sign distributed over space. Thus the assumption of negative electrons 

 made up to the present time leads to the conception of 'positive 

 electric charges distributed over a definite space. These two hypotheses 

 contain no contradiction. 



The assumption of positive electrons of parallel properties to those 

 of the negative) compels then to the conception of negative charges 

 extended through space, and thus, a- it seems to me. leads to the 

 nullification of the whole theory. Por of what use is an atomical 

 conception of electricity which cannot be subsequently worked out? 



Geophysics. — "On Frequencies of the mean </<ii/;/ cloudiness <tt 

 Batavia." By Dr. J. P. van dek Stok. 



1. Since 1880 hourly observations of the cloudiness of the sky 

 have been published by the Observatory al Batavia; if the daily 

 mean- calculated from these records are arranged in groups, a 

 frequency-table (Table I is obtained which enables us to form a 

 clear idea of the way in which the climate is affected by this highly 

 important climatological factor. 



From this Table it appears that, whilst northerly climates are 

 characterized by a great number of cases in which the sky is entirely 

 overcast or quite tree from clouds (principally in April and September), 

 these extreme values rarely occur at Batavia. 



Only once in 26 year- or in 9500 cases a serene sky lasting 

 during 24 hours has been recorded and, taken over the whole year, 

 the number of days during which the sky was entirely covered only 

 amount to 1.4% ana \ even in full West-Monsoon, to hardly more 

 than 4 %. 



Furthermore Table 1 exhibits the fact that, notwithstanding the 

 great number of records, irregularities still occur to a considerable 

 extent and the sums taken over the whole year clearly demonstrate 

 that extreme care must be taken in adding together frequency -series 

 of different kinds, which may lead to irregularities of the most pecu- 

 liar description in the curve of distribution ; and these irregularities 

 are by no means eliminated by a greater number of data. 



In order to eliminate these irregularities three natural groups have 



