320 



SCIENCE 



[N. S. Vol. XXVIII. No. 714 



to prevent penetration of air through the 

 walls, and the bougie remains filled with 

 water, although it is above the water-level in 

 the jar. Evaporation proceeds from the sur- 

 face of the bougie, water being drawn into the 

 pores to replace what has been lost. The 

 water thus removed from the cavity of the 

 bougie is in turn replaced from the reservoir, 

 and evaporation may continue as long as the 

 tube reaches the water in the latter. 



After the lapse of a time period the cork 

 stopper is loosened, slightly raised, and slipped 

 sidewise as far as the tube will permit, its 

 lower surface resting on the edge of the jar. 

 The latter is then refilled to the standard level 

 from a graduated vessel, and the amount of 

 water required to refill is the amount which 

 has been evaporated during the preceding 

 period. 



Only pure water is to be used in this in- 

 strument, for otherwise a rapid clogging of 

 the pores of the bougie ensues. Water from 

 an ordinary still is satisfactory. To prevent 

 the growth of microorganisms in and on the 

 bougie, which might clog the pores, formalde- 

 hyde may be added to the water, to make a 

 3-5 per cent, solution. This does not inter- 

 fere appreciably with the operation of the 

 instrument. The bougies, as received from 

 the factory, are, of course, not strictly uni- 

 form in porosity, and it is necessary to 

 standardize them by comparing their evapora- 

 tion rates, under uniform conditions, with 

 that from a standard bougie. The latter is 

 not used excepting for standardizing, and is 

 kept protected from dust and moisture when 

 not in use. A coefficient of correction is ob- 

 tained by standardizing, which is applied to 

 the readings obtained in actual operation. 

 If properly used the bougies will operate for 

 at least four months without appreciable 

 alteration in their coefiicients of correction. 

 It is well, however, to restandardize them at 

 the end of a season's work. 



When exposed in the open, rain may fall 

 upon the surface of the bougie, and as long as 

 this surface is wet the water movement in 

 the instrument is reversed, and water actually 

 enters the jar, at a rate determined by the 



porosity of the bougie and by the height of 

 the water-level in the jar at the time. This 

 error, is, however, very small excepting for 

 long periods of rainy weather. A correction 

 coefficient may be obtained and applied, but 

 the application must depend upon a record of 

 the duration of precipitation periods. Of 

 course the error here mentioned might be 

 avoided by placing a small screen above the 

 instrument, but such a screen would alter the 

 evaporation rate to some extent, even though 

 it were made of glass, and, since the influx of 

 heat from the sun in different localities varies 

 to a marked degree, a screen is not desirable. 



This atmometer may be read daily, weekly 

 or monthly, in fact at any convenient inter- 

 vals, the only condition being that the water- 

 supply in the reservoir must be adequate for 

 the chosen period. The tube may be length- 

 ened to several times the given length without 

 affecting the accuracy of the readings, so that 

 a large reservoir may be used for very long 

 periods. The rate of water loss from the 

 bougie may be reduced by partially coating 

 the surface of the latter with shellac, or a 

 smaller bougie may be used. 



This form of instrument has few of the 

 objectionable features possessed by the com- 

 mon open vessel atmometer. It does not at- 

 tract birds and other small animals and is 

 not subject to errors on account of their 

 visits. It is especially adapted to studies of 

 the effect of wind on evaporation, for high 

 winds do not produce any error, aS' they so 

 often do in open vessels, by the blowing of 

 liquid water from the surface and by pro- 

 ducing variation in the effectual surface itself. 

 It may be standardized so as to give results 

 in terms of depth of evaporation, as usually 

 given, by comparing it with whatever form 

 of open vessel the observer may choose as 

 standard for this purpose. It is self-record- 

 ing and self-integrating, as far as data for a 

 mean rate are concerned, and, finally, is so 

 inexpensive that several instruments may be 

 exposed at a single locality, thus decreasing 

 the chance of error. 



Burton Edward Livingston 

 Munich, Geemant 



