REPORT OF THE SECRETARY. 27 



several forms of rain gauges, we have decided to recommend the 

 general adoption of a simple cylindrical vessel of three or four inches 

 in diameter and nine inches high, the depth to be measured by 

 plunging to the bottom a small slip of soft wood on which the water 

 mark can be distinctly observed, and measuring this by the scale 

 before mentioned, the depth being recorded to the quarter of a tenth 

 of an inch. Special instructions will be given that the rain be meas- 

 ured immediately after the fall or before any sensible evaporation 

 has taken place. We have adopted the simple cylinder of uniform 

 diameter as being less liable to errors of observation than any other 

 form. 



Although the separate observations are not read with the same 

 minuteness as in the case of gauges in which the depth of rain is mag- 

 nified by a receiver of less diameter than the orifice of the gauge, yet 

 the average we find from experience in the case of a long series givqs 

 equally reliable results with those in which instruments of apparently 

 greater precision are employed. 



After the completion of the rain tables, our computers will com- 

 mence the discussion of the temperature of the North American con- 

 tinent. Were our funds suflScient, we should be glad to include in 

 the investigation all the observations made on this continent during 

 the various expeditions to the arctic regions, undertaken by the Brit- 

 ish government, few of which have, as yet, received that thorough 

 examination necessary to obtain from them the general truths which 

 constitute real contributions to science. 



We have mentioned in previous reports that the meteorological sys-' 

 tem of the U. S. army was about to be reorganized under the Surgeon 

 General. This work has been carried on during the year, and a series 

 of standard instruments has been constructed by James Green, of 

 New York, under the direction of Dr. Craig, for distribution to the 

 various posts. Fifty barometers and one hundred and fifty thermom- 

 eters and rain gauges will be substituted for those now in use. 



The number of Smithsonian observers during the year 18G7 was 

 385, and this will probably be increased during the year 18G8, 

 so that a more rapid and accurate accumulation of data relative to 

 the meteorology of this country will be obtained than at any time 

 heretofore. 



The Department of Agriculture has continued during the past year 

 to publish the monthly bulletin of meteorological observations, which 

 is still received with much interest by farmers, as well as by meteo- 

 rological observers. The preparation of the meteorological notbs for 



