: 128 
corder, and a continued record made of the temperatures to which 
the roots of plants are subjected. The amount of liquid in the 
capillary tube and corrugated chambers is so small in comparison 
with that in the bulb, that the variations in volume of this por- 
tion of the confined liquid make no practical differences in the 
readings. Thus no matter how warm the air might become, or 
even if the sun shone directly on the capillary tube the instru- 
ment would show the temperature of the bulb to within a very 
small fraction of a degree, the error being so small as to be neg- 
ligible in all ordinary experiments. This is further illustrated by 
the fact that the flame of a bunsen burner may be applied to 
this tube without apparent effect on the temperature shown by the 
indicator. 
But one of the Hallock thermographs has been constructed, 
and it has been installed as described below in a location in the 
Meteorological station in Garden. Rain-gauge in corner of enclosure. 
Fic. 20, 
The bulb of the Hallock thermograph is buried in the middle of the enclosure at a 
The instrument shelter at the right contains the recording and 
pth of 30 ¢ 
registering speaetl us. 
nurseries of the Garden, to which place all of the meteorological 
apparatus has been removed. The recording portion of the in- 
strument, together with an air thermograph, has been placed in 
an instrument shelter of the United ees en Bureau pattern. 
(See Fig. 20.) The capillary tube! dedd through 
the floor of the shelter to the surface of the ground, being protected 
