6 DYNAMIC METEOROLOGY AND HYDROGRAPHY. 



Thus a tolerably satisfactory simultaneity can be realized even for the observa- 

 tions from the higher strata. But still the principle of simultaneity is not carried 

 through universally, not even for the observations at the ground, where its realization 

 should not cause any real difficulty. Thus departures by far exceeding the half -hour 

 limit exist still in the European net of daily observations. Fortunately in the 

 United States the principle of simultaneity is completely carried through for the 

 whole net of stations. This circumstance, in connection with the complete homo- 

 geneity of the observations, all being obtained from self-recording instruments of 

 the same construction and treated according to the same rules, make these observa- 

 tions the best which we have had at our disposal for the study of the conditions of 

 the atmosphere near the ground. 



Passing to the distribution in space of the points of observation, we must 

 distinguish the points of observation near the ground from those in the free atmos- 

 phere. As to the investigation of the lowest atmospheric sheet, the greater nets of 

 observation, as that of Europe, of the United States, or of India, may be said to be 

 satisfactory, exceptions being made for certain specially difficult regions, for instance 

 the western mountainous parts of the United States. For practical reasons the net 

 of stations is here less close, while the space- variations of meteorological elements are 

 stronger than in the flat land. For the most variable element, wind, this has caused 

 us great difficulties. 



In the free space fixed points of observation can not be maintained, and would 

 not, unless they could be kept up in great number, be of appreciable use; for the 

 lengths which can be used as line-differentials in vertical direction are much smaller 

 than those which can be used in horizontal direction. But on account of the relative 

 slowness of the variations in time and the rapidity with which meteorological 

 ascents can be performed, we can get continuous records along vertical lines, repre- 

 senting approximately the instantaneous state of things along these lines. 



As the variation of meteorological elements in horizontal direction is necessarily 

 much smaller in the free atmosphere than near the ground, where the local influences 

 of topography come in, it will not be necessary to provide all stations at the ground 

 with the implements for meteorological ascents. But only experience can show how 

 close the net of aerological stations should be. Further, it will not be required to 

 give all aerological stations equally complete equipment, for the scalar elements 

 have much less pronounced space-variations than the vector-element, velocity. 

 As air- velocity is also much easier to observe, thanks to the method of pilot-balloons, 

 it will be rational and economical to erect two classes of aerological stations, com- 

 plete aerological stations and pilot-balloon stations. How close the net of each 

 kind should be, will be evident by and by from the synoptical maps drawn by use 

 of the ascents. The erection of aerological stations, including pilot-balloon stations 

 in great numbers, will be of special importance in mountainous regions, where the 

 effectivity of the common stations is so limited on account of the local irregularities. 



The last and most delicate question is that of the determination of a suitable 

 time-differential separating the epochs of observation. Inasmuch as continuity in 

 time is realized in as great extent as possible by providing the stations at the ground 



