A Defcription of an Hydrojlatic Lamp, 469 



would be readily underflood that it would float when oil was introduced at the lip. And 

 as the floatage would be regulated or confined by the horizontal axis, there were two very 

 curious properties obtained ; that the bulk of the folid above the vefl"el would always be 

 equal to that of the oil itfelf, and the oil would remain at the fame height till it was all 

 burned. It may perhaps afford fome amiifement to the reader to deduce thefe properties 

 from the {truflure of the machine, which may be done with very little confideration. 



Dr. Hooke's contrivance, neverthelefs, though worthy of great admiration, appears to be 

 pradlically nothing more than an improvem.ent on the fhallow difh lamp. For at the com- 

 mencement the light does not illuminate much more than one hemifphere of the furround- 

 ingfpace; and when the float has defcended, towards the end of the confumption, the illu- 

 minated part is upwards of three-fourths of that fpace. 



A much completer folution of the pra£tical problem of illuminating the greateft polli- 

 ble fpace, while the fall of the oil is rendered the leaft poflTible, was given to the public by 

 Mr. Keir, of Kentifli Town, in the year 1787, for which he obtained a patent. It feems 

 probable that the advantages of this contrivance being directed to an objeft fo much lefs 

 ftriking than t^iat of the lamp of Argand, which was then new in the mind of the public, 

 might, together with the confined . exhibition of an article referved to a few hands, have 

 prevented it from becoming generally known ; and as it is probably unknown to many of 

 my philofophical readers,. I have thought an account of its confl;ru£lion would be accept- 

 able. 



In plate XX are exhibited a view and feflion of one of thefe lamps. The feGion only 

 requires to be explained. From the flender figure of the vafe it is evident that the flame is 

 permitted to throw its light in all direftions downwards and upwards, nearly in the fam? 

 manner as a candle. The interior parts is divided into feveral compartments by the dia- 

 phragms at F and C. The fpace A A, above F, is open to the atmofphere ; but the fpace 

 B B, beneath C, is clofe. A tube F G proceeds from the fpace A A to the fpace B B, fo as 

 to reach nearly to the bottom at G : and another tube C D, proceeds from B B upwards, 

 through A A, without communicating with this laft fpace, and is enlarged at the upper 

 part, fo as to receive a wick with the apparatus of Argand, or any other. A folution of 

 fea fait, or the mother water of fait, being firft poured in by meafure at E, flows down the 

 tube into B B, and fills that fpace. A like meafure of oil is next poured, which alfo de- 

 fcends into B B, and forces the denfe faline liquor upwards, through G F, into the fpace 

 A A. The fpecific gravity of this laft is adjufted by dilution, fo that when the fpace A A Js 

 properly filled the oil fhall ftand in equilibrio at the requifite height near E; that is to fay, 

 the furfaces in A, and at E, are elevated above the lower orifice at G in the inverfe pro- 

 portion of the fpecific gravities. This proportion is ufually about 3 to 4. So that if any 

 part of the oil be taken away from E by combuftion, or othcrwlfe, there will be a fubfi- 

 dence of the heavy fluid in A A to preferve the equilibrium ; and during the whole fubfi- 

 dence in A A there will be a correfpondent dcpreffioa of the upper furface of oil, near E, 

 which will be meafured by four third parts of the firfl elevation of the denfe fluid above tiie 

 Vol. III.— January 1800. 3P partition. 



