86 



SCIENCE 



[N. S. Vol. LII. No. 1334 



ments of water-loss by evaporation. The 

 reading may be nil or even negative (entrance 

 of water into tbe reservoir), in spite of the 

 actual evaporation of significant amounts of 

 water from the instrument during the periods 

 between showers. 



It is not feasible to correct for these errors 

 of water absorption, but the difiiculty has 

 long been practically overcome by the em- 

 ployment of a rain-correcting, or non-absorb- 

 ing, mounting for this kind of atmometer 

 when operated in the open during rainy 

 weather.2 ^]l ^j^g non-absorbing mountings 

 thus far suggested depend upon a mercury 

 valve that allows water to pass through the 

 tube freely in the upward direction, but 

 practically prevents movement downward. As 

 soon as liquid water is deposited on the 

 exposed porous surface the surface automatic- 

 ally becomes virtually impervious to water 

 from without, and the precipitation water 

 rims off from the instrument as though it 

 were glazed. When the rain ceases the valve 

 reverses and evaporation soon begins to be 

 registered as water-loss from the reservoir. 

 Various types of mercury-valve mounting 

 have been described, but the Shive form has 

 been most generally used. All these non- 

 absorbing mountings are relatively expensive, 

 and the least expensive one (Johnston's) in- 

 volves the use of rubber and requires special 

 care in the installing of the instrument. 



A much simpler form of mounting than 

 any hitherto sugested has recently been tested 

 in the Laboratory of Plant Physiology of the 

 Johns Hopkins University. The purpose of 

 this paper is to place the new modification in 

 the hands of those who are interested in 

 atmometry, so as to save them the expense 

 and troubles of the more complicated mount- 

 ings. 



2 Livingston, B. E., "A Eain-correeting Atmom- 

 eter for Eeological Instrumentation," Plant 

 World, 13: 79-82, 1910. Harvey, E. M., "The Ac- 

 tion of tlie Rain-correcting Atmometer," Plant 

 World, 16: 89-93, 1913. Shive, J. W., "An Im- 

 proved Non-absorbing Porous-cup Atmometer," 

 Plant World, 18: 7-10, 1915. Johnston, E. S., "A 

 Simple Non-absorbing Atmometer Mounting," 

 Plant World, 21: 257-260, 1918. 



The new mounting is very simple. The 

 porous-porcelain piece is mounted in the 

 usual way, by means of a rubber stopper, on 

 the upper end of a glass tube of suitable 

 length and having a bore of about 6 or 7 

 mm. This tube bears a second rubber (or 

 cork) stopper somewhat below the first, which 

 fits the mouth of the reservoir bottle and 

 closes it completely as far as entrance of rain- 

 water is concerned. The reservoir stopper is 

 not slotted to allow air entrance to the reser- 

 voir, but access of air to the interior is 

 allowed through a short, inverted-U-shaped 

 glass tube, one arm of which is longer and 

 penetrates just through the reservoir stopper 

 from without, while the other arm is shorter, 

 is directed downward and terminates a few 

 millimeters above the upper surface of the 

 stopper. This U-tube may be very small and 

 its end may be loosely plugged with glass 

 wool to exclude insects, etc. A water-proof 

 apron over the top of the reservoir may be 

 employed (Livingston, 1908), or other devices 

 to allow air entrance and to exclude rain 

 water may be used. 



Thus far we have an obsorMng mounting, 

 suitable only for indoor operation or for 

 periods without precipitation. But a very 

 simple and efficient mercury valve is in- 

 serted in the upper end of the straight 

 tube, as follows. A tightly rolled plug of 

 glass wool (about 1 cm. long) is inserted in 

 the upper end of the tube, the outer end of 

 the plug is cut off so as to have a flat surface, 

 and it is pushed into the tube imtil its upper 

 end is about 2 cm. from the top of tlie tube. 

 Next, a small amount of mercury is placed 

 in the tube above the plug (the mercury 

 column being 5-8 mm. high) and another 

 plug of like nature is inserted above the 

 mercury. The mercury is imprisoned between 

 the two plugs and can not escape, in what- 

 ever position the tube is placed. 



To install the instrument, the tube is in- 

 verted and the end bearing the valve is in- 

 serted in distilled water while suction is ap- 

 plied at the other end. Water enters freely 

 through the valve and the tube is nearly 

 filled in this way. Then the porous-porcelain 



