CONSIDERATIONS ON THE OBJECT AND METHODS. 7 



with self-recording instruments, the question will be reduced to that of a suitable 

 interval between the successive aerological soundings. As time- variations of the 

 meteorological elements have the same rapidity near the ground as in the free air, 

 this question can be answered by examination of charts for the ground concerning 

 the element which has the most rapid time- variations, namely, velocity. According 

 to our preliminary experience regarding these charts (see Chapters XII and XIII) 

 it seems reasonable to try time-differentials of three hours for this element, while 

 differentials of double the length may be used for the other elements. 



Observations of the completeness thus required can not be kept up continuously. 

 It will be necessary to organize special periods of investigation extended for each 

 time over a series of days. An effective organization of such a period would be this : 



During the whole period continuous observations or observations for every 

 hour of Greenwich time are kept up at all stations at the ground. 



For every third hour of Greenwich time ascents are made from the pilot- 

 balloon stations. 



For every sixth hour of Greenwich time ascents are made from the complete 

 aerological stations. 



94. Remarks on Hydrographic Observations. Oceanographic observations are 

 not yet organized systematically. But the general principles for their organization 

 will be the same as for the meteorological observations. Hydrographic expedi- 

 tions going out occasionally can only contribute to the knowledge of the average 

 state, i. e. , to the climatology of the sea. But the final aim must be that of investiga- 

 ting the actual states and their variations. The organization must then be governed 

 by the principle of simultaneity. The investigations will have to be performed not 

 by one luxuriously fitted ship, passing months or years at sea, but by the cooperation 

 of small ships going out simultaneously. 



The demands regarding the degree of simultaneity and the intervals between 

 the epochs of observation will depend upon the rapidity of the changes. There are 

 indications both for rapid changes (among which the tidal phenomena in the deeper 

 strata will play an important part) as well as for slow seasonal changes and changes 

 from year to year. The problem will be to organize observations so as to separate 

 from each other the changes of different rapidity and to investigate them as much as 

 possible independently of each other. But a serious discussion on the suitable 

 method of organization will only be possible by and by, as our knowledge of the 

 oceanic phenomena advances. 



95. The Problem of Diagnosis. The observations being given, the diagnosis 

 will consist in working out continuous synoptical representations of the field of each 

 element. This involves first the choice of proper methods of representing each field 

 synoptically. This choice being made, methods for passing from the observations to 

 the synoptic representation must be worked out. These diagnostic methods must 

 take into consideration not only the observations themselves, but also all intrinsic 

 relations existing between observed quantities and quantities to be represented. It 

 is due to these intrinsic relations that we are able to work out relatively complete 





