324 



METEOKOLOGY AND ALLIED SUBJECTS. 



ing figures, showing the average visibility estimated on an arbitrary 

 scale, in which = unusually clear, and 5 = densest fog or haze; 



On the average, Dr. Hamberg finds that the greatest clearness occurs 

 with winds between N. and W., and the greatest obscurity with winds 

 between S. and W. ; these cases, of course, correspond to positions of 

 the barometric minima, respectively, northeast and northwest of Upsala. 

 A comparison of these observations witli the relative humidity of the 

 air shows that in general transparency diminishes with an increase of 

 moisture, and especially is this the case during the colder weather. A 

 direct dependence upon cloudiness is not evident, and the author con- 

 cludes that the i)artial condensation of the aqueous vapor in the atmos- 

 phere and the presence of dust or smoke are the fundamental causes of 

 the variations in transparency. Of the origin of the material compos- 

 ing the day fog nothing definite is known. As a means of foretelling 

 rainy weather the transparency of the air is frequently appealed to. 

 Hamberg finds that the quantity of rainfall increases directly as the 

 transparency diminishes. {Z. 0. G. il/., XYI, 1881, p. 457.) 



XII. — MISCELLANEOUS MATTERS. 



(a) Periodicities. — S. A. Hill has investigated a decennial period in the 

 annual variation of temperature and pressure in India. He states that 

 the heat of the summer of 1878, especially during the first twenty-one 

 days of June, was so great in Northern India that nothing like it had 

 ever before been experienced. This excessive summer heat was pre- 

 ceded and followed by winters of unusual cold. He has therefore col- 

 lected all available observations bearing upon the question of a possible 

 periodicity in temperature and pressure. The temperature ai)i)eai's to 

 follow the variations in the sun spots. He finds that at Calcutta the 

 greatest variations in pressure occurred in the years 1845, 1857, 1806, 

 and 187G, and the least variations in the years 1840, 1847, 1801, 1872, 

 and 1874, and these dates are only slightly modified if we take the means 

 of groups of three years each. The above dates agree closely with the 

 minima and maxima of the sun spots. An exi)lanation of this coinci- 

 dence may be about as follows : The general distribution of winds on 



