INVESTIGATIONS RELATIVE TO ILLUMINATING MATERIALS. 507 



order charged with mineral and with hird oil, all the experiments we 

 have yet made on this point tend to the conclusion that in smaller lamps 

 with the more volatile oils a greater photometric power is obtained than 

 with the same lamp Avhen charged with lard-oil ; but with the larger 

 lamps the reverse is the case, the lard giving greater power burned in 

 these lamps than the mineral oil. 



An unexpected difficulty arose in the course of the investigations for 

 the introduction of mineral oils on account of the form of the flame. 

 While a lamp with a constricted chimney, like that used in the German 

 student-lamp, gave the greatest photometrical power, it was found that 

 the shape of the flame did not correspond with the arrangement of the 

 lens-apparatus, a large portion of the light being thrown upward toward 

 the sky and another toward the earth. It was only after a series of 

 trials with chimneys of different forms and button-deflectors that a flame 

 of the best shape was obtained. To compare these flames in actual use, 

 they were placed in succession in a lighthouse, with a lens of tlie fourth 

 order, and the photometrical power determined at different distances, 

 from a mile to ten miles in extent, by interposing between the eye and 

 the light a series of thin colored glasses, until the light was totally ex- 

 tinguished. It was found in these experiments that some of the flames 

 that had an appearance of greater brilliancy near by failed to produce 

 comparatively the same effect at a greater distance. Having settled 

 upon the form of the flame to be used in lamps of the lower onlers, ar- 

 rangements have been made for the introduction of mineral oils into all 

 the stations in the third district, at which lights of the fourth and 

 smaller orders are at present in use. The substitution of mineral for 

 lard oil however is a matter of no small difficulty, and requires to be 

 made with great precaution. An entire change in all the lamps is re- 

 quired ; the several parts of the apparatus which in the case of lard- 

 oil lamps were united by soft solder must now be joined with spelter. 



The importance of this was evinced by an accident which hapi)ened 

 in the photometric room in the case of a lamp of the fourth order under 

 trial; the heat unsoldered an air-tube and let down the oil from the res- 

 ervoir on the flame, which produced so fierce a combustion that it would 

 have set fire to the building had it not been of fire-proof materials. 



The gradual introduction however of mineral oil will be made as 

 rapidly as experience indicates the best and safest mode of employing 

 it. It has already been adopted in the smaller lamps for lighting the 

 Mississippi and its principal tributaries. The substitution however is 

 not on account of the superior quality of this oil in comparison with 

 lard, since we think the latter as an illuminating material is inferior to 

 no other at present in use, but simply on account of the comparative 

 cost of the two materials. 



The comparative cost of the two materials will be definitely ascer- 

 tained after we have determined the best form of lamps to be used. 

 Experiments thus far have been principally confined to the lower orders 

 of lamps. 



