72 Exper'imetUs for the Improwment of Cuiidlis, 



enough to abforb the tallow, and at the fame time fufficiently flexible to bend on one 

 fide. 



The mofl promiCng fpeculation refpc£ling this mod ufeful article, feems to direft itfelf 

 to the cup which contains the melted tallow. The imperfeifliou of this part has already 

 been noticed, namely, that it breaks Aovm by fufioii, and fullers its fluid contents to efcape. 

 The Chincfc have a kind of candle about half an inch in diameter, which, in the harbour 

 of Canton, is called a kkhock ; but whether the name be Chinefe, or che corruption of fome 

 European word, I am ignorant. The wick is of cotton, wrapped round a fmall ftick or 

 match of the bamboo cane. The body of the candle is white tallow ; but the external 

 part to the thicknefs of perhaps one-thirtieth of an inch, confifts of a ,waxy matter 

 coloured red. This covering gives a confiderable degree of folidity to the candle, and pre- 

 vents its guttering, becaufe lefs fufible than the tallow itfelf, I did not obfervc that the 

 ftick in the middle was either advantageous or the contrary ; and, as I now write from the 

 recoUetlion of this obje£l at fo remote a period as twenty-five years ago, I can only con- 

 je£lure that it might be of advantage in throwing up a lefs quantity of oil into the flame than 

 would have been conveyed by a wick of cotton fufficiently ftout to have occupied its place 

 unfupported in the axis of the candle. 



Many years ago I made a candle in imitation of the lobchmch. The expedient to which 

 I had recourfe confifled in adapting the wick in the ufual pewter mould: wax was then 

 poured in, and Immediately afterwards poured out : the film of wax which adhered to the 

 inner furface of the mould foon became cool ; and <lie candle was completed by filling 

 the mould with tallow. When it was drawn out, it was found to be cracked longitudi- 

 nally on its furface, which I attributed to the contradion of the wax, by cooling, being 

 greater than that of the tallow. At prefcnt I think it equally probable that the cracking 

 miglit have been occafioned by too fudden cooling of the wax before the tallow was poured 

 in ; but other avocations prevented the experiments from being varied and repeated. It is 

 probable tliat the Chinefe external coating may not be formed of pure hard bleached wax. 



But the moft decifive remedy for the imperfe£lion of this cheaped, and in other refpecl* 

 betl material for candles, would undoubtedly be to diminifh its fufibility. Various fub- 

 ftances may be combined with tallow, either in the diredl or indire(5l method. In the 

 latter way, by the decompofition of foap, a number of experiments were made by Ber- 

 thoUet, of which an account is inferted in the Memoirs of tlie Academy at Paris for the 

 year 1 780, and copied into the 26th volume of the Journal de Phy fique. None of thefe point 

 dirtclly to the pvefent obje£l ; befides which, it is probable that the foap made ufe of 

 by that eminent chemift was formed not of tallow, but oil. I am not aware of any 

 regular fcries of experiments concerning the mutual a£lion of fat oils and other chemical 

 agents, more cfpccially fuch as may be direfled to this important object of diniinifliing its 

 folubility ; for which reafon I Ihall mention a few experiments made with this view. 



I. Tallow was melted in a fmall filver veflcl. Solid tallow finks in the fluid, and diflblves 

 without any remarkable appearance. 2. Gum fandarach in tears was not diflblved, but 

 emitted bubbles, fwcllcd up, became brown, emitted fumes, and became crifp or friable. 

 No folution nor improvement of the tallow. 3. Shell-lac fwellcd up with bubbles, and 

 was more pcrfeftly fufcd than the gum fandarach in the former experiment. When the 

 tallow was poured 00", it was thought to congeal ratlicr more fpeedily. The lac did not 



appear 



