2448 Journal of Applied Microscopy 



greatly reduces the arc. Rather heavy silver contacts were found to be efficient. 

 To avoid any danger of fire, we placed the cup of oil in a tin box which also 

 served as a stand for the slate base. 



The thermostat may need a more careful description. Two holes were made 

 through a rubber cork. Through the one in the middle a glass with a small 

 bore was inserted. A copper wire was placed through the other opening near 

 the side of the cork. A test tube was filled with mercury and the cork was 

 inserted. Enough mercury was poured into the glass tube to raise the level of 

 the liquid a little way above the cork. The test tube was then fastened to a 

 frame made of insulating material. A copper post firmly fastened to this frame 

 projected some distance above the top of the glass tube. Turning at right 

 angles it held a burr at its free end. A brass thumb screw to the end of which 

 was soldered a soft iron wire screwed up or down through the burr. 



To regulate the thermostat all that is necessary is to raise the end of the 

 iron wire high above the mercury. When the bath has reached the desired tem- 

 perature, screw the wire down until it just touches the mercury. The lamp (G) 

 was placed in the circuit as a resistance to keep the magnet (E) from becoming 

 overheated. In selecting the wire for the thermostat two properties were taken 

 into consideration. The wire must not readily amalgamate with the mercury or 

 it will be eaten away, thus destroying its accuracy. If the metal has a strong 

 adhesion for mercury the top of the mercury column will be drawn out into 

 quite a long neck before it breaks the circuit. A weakness of this sort greatly 

 affects the variation of the temperature. We have not experimented very exten- 

 sively, but we find platinum to be very poorly adapted on account of adhesion. 

 Soft iron appears to work very nicely. The thermostat presents one especial 

 weakness. When the circuit is broken a spark is formed which gradually uses 

 up the iron wire. When immersed in light oil this difficulty is obviated, but 

 each spark carbonizes a little oil. After using for several weeks, the tube is 

 blackened to necessitate the removal of the wire and the cleansing of the tube. 

 If some one can find an oil that will not readily carbonize, it will certainly add 

 greatly to the value of the thermostat. 



I see no reason why this device cannot be adapted to larger baths, incuba- 

 tors, and ovens. All that would be necessary would be to substitute a heating 

 coil for the electric lamp. If a constant temperature with less variation is needed 

 it can be obtained in either of two ways. Two heating coils or lamps can be 

 used ; the one which should be insufficient to heat the bath in operation all the 

 time, the other regulated by the apparatus described merely keeping the temper- 

 ature up to the proper point. Enlarging the mercury cistern of the thermostat 

 will also reduce the variation. Jesse J. Myers. 



Michigan Agricultural College. 



