56 
BULLETIN OF THE BUREAU OF FISHERIES. 
The Schmidt-Vossberg thermometer, which was used on Crater Lake September 
5, 1913, and on Lake Chelan, was graduated to 0.1° C. It was not standardized 
but was checked at the New Mexico School of Mines especially at 4° C. and the 
correction applied. 
BOTTOM TEMPERATURES. 
In lakes not over 100 m. deep it is comparatively easy to tell when the ther- 
mometer strikes bottom. It is then raised 0.25 to 0.5 m., allowed to become 
constant, and tripped. In the deeper lakes the thermometer did not stretch the 
line as much as the sounding weight and sampler, so it was difficult to tell when 
the bottom was reached by depth, and it could not be detected by feeling. 
To overcome this difficulty, a stone or weight was fastened by a loop to the hook 
on the lower end of the trip thermometer, so that it would be 0.5 m. below the 
thermometer (which gave room for the thermometer to trip without hitting the 
stone or bottom) . With this it was easy to locate the bottom, and when the ther- 
mometer was tripped the stone unhooked. 
During August, 1911, a set of temperatures was taken on Hayden Lake, Idaho, 
using a common glass-rod thermometer inserted through the rubber stopper in the 
upper end of the water sampler. To a depth of 56 m. the results agreed within 
0.5° C. with those taken with the trip thermometer. The sampler must be hauled 
up quickly and read at once. It is mentioned here as a simple, rough method of 
determining temperatures of shallow to medium depth lakes. 
SAMPLES OF WATER. 
APPARATUS USED IN OBTAINING SAMPLES. 
The original plan was to obtain the samples of water with a pump and hose 
similar to that used on the Wisconsin lakes (Birge and Juday, 1911). The outfit 
consisted of a rotary pump, 90 m. (300 feet) of J-inch hose, and a calibrated brass 
chain of the same length. This outfit proved to be very hard to handle on lakes 
50 to 60 m. deep, and it could not be used on the very deep lakes. The first sets 
of samples on Hayden and Coeur d’Alene lakes were taken with this outfit, after 
which it was abandoned and all other samples were taken with a sampler. 
THE SAMPLER. 
The sampler used on Lake Pend Oreille and Upper Priest Lake in 1911 was 
loaned by the Wisconsin Geological and Natural History Survey. The line parted 
and the sampler was lost in Priest Lake. The sampler used during the remainder 
of the summer was designed and built in Spokane. Another sampler was built at 
the New Mexico School of Mines. It is similar to the one built in Spokane, except 
that it is a little shorter and is screwed together, whereas the Spokane model was 
soldered. A description of it follows, as it is thought to be the simplest and cheapest 
apparatus to be had for taking samples in deep water. 
The sampli n g apparatus (fig. 3) consists of a 16-inch piece of 2^-inch brass 
tubing (A) that is closed by rubber stoppers ( E ) at each end. The lower one of 
these stoppers is attached to an |-inch brass pipe (<7) that passes through guides in 
the tube ( A ). The upper stopper is attached to a piece of brass tubing (D) that 
