228 PHENOMENA CONNECTED V/ITH CLOUDY CONDENSATION. 



fore require to be taken, aud a correction made wliicli it niij^lit be 

 difficult to carry out in practice. 



At present the best plan of graduating seems to be to note the depth 

 of the blue produced, regardless of the amount of expansion required 

 to give it, and use only this quantity as an index. With few particles 

 the color is pale, and as the particles increase in number the color 

 increases in depth. Perhaps some addition might be made to the 

 instrument for estimating more accurately the depth of color than can 

 be done mentally. This might be done either by means of colored 

 glasses of different depths for comparison, or in some other Avay. 



A few comparative observations have been made with the Jxoni.scopc 

 and the dust-counter in the impure air of a room, AVhile the number 

 of particles was counted by means of the dust-counter, the depth of blue 

 given by the loniscope was noted. A metal tube was fitted u}) verti- 

 cally in the room in such a way that it could be raised to any desired 

 height into the impure air near the ceiling, so that su])plies of air of dif- 

 ferent degrees of impurity might be obtained. To produce the inq)urity 

 the gas was lit and kept burning during the experiments. The air was 

 drawn down through the pipe by means of the air pump of the T^oni- 

 sco_[)e, ami it passed through the measuring apparatus of the dust- 

 counter on its way to the l-onlscopr. The indications of the two instru- 

 ments were taken, and are entered in the following table: 



It is probable that the higher numbers are too low, as the measure of 

 the dust-counter has a. cai)acity of only 10 cubic millimeters. With so 

 small a measure it is probable that a good many of the particles are 

 lost. When making a sanitary ins])cction, the air outside, or wherever 

 the supi)ly was drawn from, would be tested first, and the depth of color 

 which it gave would be noted. Any increase from that depth would 

 indicate that the air was being jiolluted, and the amount of increase in 

 the depth of color would indicate the amount of increase of pollution 

 Slight color can be traced, though the number of particles be less than 

 80,000 per cubic centimeter, but the color is not very decided, the con- 

 densation producing principally a darkening effect. It should be noted 

 that the above table refers to a koniscope with a test tube 50 centi- 

 meters long. An instrument with a tube 1 meter long would be doubly 

 sensitive and would show color with fewer particles. 



II 



