404 Professor Draper's Description of the Tithonometer. 



Let A B, fig. 4, be an Argand oil lamp of which the wick 

 is C. Over the wick, at a distance of half an inch or there- 

 abouts, place a plate of thin sheet copper, three inches in 

 diameter, perforated in its centre with a circular hole of the 

 same diameter as the wick, and concentric therewith. This 

 piece of copper is represented at dd; it should have some 

 contrivance for raising or depressing it through a small space, 

 the proper height being determined by trial. On this plate, 

 the glass cylinder e, an inch and three-quarters in diameter 

 and eight or ten inches long, rests. 



When the lamp is lighted, provided the distance between 

 the plate dd and the top of the wick is properly adjusted, on 

 putting on the glass cylinder the flame instantly assumes an 

 intense whiteness ; by raising the wick it may be elongated to 

 six inches or more, and becomes exceedingly brilliant. Lamps 

 constructed on these principles may be purchased in the shops. 

 I have, however, contented myself with using a common Ar- 

 gand study-lamp, supporting the perforated plate d d at a 

 proper altitude by a retort stand. It will be easily understood 

 that the great increase of light arises from the circumstance 

 that the flame is drawn violently through the aperture in the 

 plate by the current established in the cylinder. 



As much radiant heat is emitted by this flame, in order to 

 diminish its action, and also to increase the tithonic effect, I 

 adopt the following arrangement. Let A B, fig. 4, be the 

 lamp ; the rays emitted by it are received on a convex lens D, 

 four inches and three-quarters in diameter, that which I use 

 being the large lens of a lucernal microscope. This, placed 

 at a distance of twenty-one inches from the lamp, gives an 

 image of the flame at a distance of thirteen inches, which is 

 received on the sentient tube of the tithonometer F; between 

 the tithonometer and the lens there is a screen E. 



Things being thus arranged, and the lamp lighted so as to 

 give a flame about three inches and a half long, the experi- 

 ments may be proceeded with. It is convenient always to 

 work with the flame at a constant height, which may be de- 

 termined by a mark on the glass cylinder. At a given in- 

 stant, by a seconds watch, the screen E is removed, and im- 

 mediately the tithonometer begins to descend. When the 

 first minute is elapsed the position on the scale is read off' and 

 registered ; at the close of the second minute the same is done, 

 and so on with the third, &c. And now, if those numbers be 

 compared, casting aside the first, they will be found equal to 

 one another, as the following table of experiments, made at 

 different times and with different instruments, shows : — 



