THE ANALYST. 183 ON THE RATIO OF EXPANSION BY HEAT OF BUTTER FAT, LARD FAT, AND BUTTER SUBSTITUTES. BY G. W. WIGXER, F.C.S. I HAVE made some accurate deterrniiintioiis of the ratio of expansion of fats, especially butter fat and the fats used for adulterating butter. These determinatioiis will be of service, since, by tlieir use, the specific gravity of melted fats may be taken at temperatures varying within reasonable limits from tlic looo F.usually adopted, and the results so obtained may be directly compared with the specific gravity at looo. They will, also, have the advantage of rendering tlie use of tlie specific gravity bubbles, which I proposed some years ago::: for this purpose, more convehieut. #t See ANALYST, vole i., 1). 146.184 THE ANALYST. I have tested the ratio of expansion of butter fat for every 50 betmeen looo Falir.and Tlie determinations have been made independently by three different processes. 1st. By direct weighing of the melted fat in a specific gravity bottle, wit11 thermometer stopper. At tlie higher temperatures this method is certainly open to greater errors than are desirable, on account of the low conductivity of the melted fat for heat, which reiiders it a long task to secure an absolute temperature throughout the bottle.Tliis process would, lxoba>bly, give the most accurate results, if it were not for tlie tendelicy of the different €ah to crystallize, and, apparently, to Borne extent to separate when enclosed in the bulb, without the possibility of agitating them. This latter process gives results slightly inferior in accuracy to the other processes, when they are carried out in the most careful way ; but it is yet so simple and certain in its results, that it is impossible, with ordinary care, to incur any error which is of importance, when tlie process is used for analytical purposes, I have not corrected thc figures for tlie cubical expansion of glass, because ',he results are primarily intencled for the use of analysts who will mcigli tlie fats, or measure them in glass vessels, as I have done.1 s t . E.rpnnsion 01 Pure Butter Fat.-100 volumes of pure butter fat, at looo F., expaiid to 1047.2 volumes, at 2120 F. This result is the average of seven closely agreeing determinations, on four different samples of butter.This gives an average expansion of 00434 per degree F., or e0780 per degree Centigrade. The ratio of expaiisioii is not absolutely uniform throughout, but differs between 150° and 100° I!. From 100° I?. to 150° F. it is smsibly in accordance with the average ; from 150" F. to 190° F. it incrcascs sliglitly ; and from 190@ F. to 210° F, it appears to expand at its former rate of *0434".This change in the ratio of expansion is morc clearly shown by tlie lithographed diagraiu wliicli accompanies this paper. At first I was inclined t o attribute tliis cliangc to a partial decomposition of some of tlie constituents of the fat, but, on examining a sample of lard in the same way, I fouiicl that an almost identical change took place, and at nearly the same temperatures.Since lard has already been melted at much higher temperatures, and is practically free from tlie soluble and volatile fatty acids, tliis explanation seems donbtful. An artificial butter-" butterine," also showed a similar abnormal rate of expansion, but I cannot speak so exactly as to the temperatures at which the ratio changed. 2nd. Exyamion of Lard and Bzctteo.ine.-Tlie average expansion of lard and butterine (animal) are almost identical.The average ratio for lard, as deduced from tlie experiments, is *0420 per degree F. Tlie difference from the butter ratio is so small that it is of no importance for analytical purposes, except for wide ranges of temperature. It will bc more coiivenieiit for the use of these figures in the detection of nclulteratioii, to reduce these results to the ordinary figures of actual density, as compared with water at 100 O , Actual density, strictly speaking, is weight, as compared with the weight of the same bulk of water at the same temperature, and it was in this sense tliat lluter first used the term in butter analysis, but it is not convenient to 212O.2nd. By means of thermometer tubes with large bulbs.3rd. By means of specific gravity bubbles of various weights.RATIO OF EXPANSION OF FATS (3. W. WIGNER, ‘‘ The Analy6t,” OCTOBER, 1879. I [ I 1 ______l___-l_ .___- ----------- ----I---- -_ I ~~ i i I I 1 , I , I I 1 I I I I 1 1 I i I I I I I . I 1 1 I l l 1 I 1 I I I I t ’ . I I . . . . . < . . . I . , , , I . , . , . . : . . . ! , . . , ! , , , , , , ., , , ! ! ! , ! ,.!!H,!! . . . _ _ ! . . I . A Aotud Density of Lard at3 oombined with water at looo. B P P ,, Butter of 911O Actud Deneity. c *P ,, Mixed Butters of 912.6 Actud Density. D Curve of Expansion of Mhter. (The figures of weight do not apply to this curve.) At??&&i?lW* 860 a70 1890 ,800 ‘910 920 0’THE ANALYST. 185 follow this rule at temperatures above 100 O, because whereas water expands more rapidly with increase of temperature, fats do not do so.I therefore compare all the figures obtained with water at looo as the standard. Assuming the limit of 911 as the lowest average actual density of a genuine butter at 1000 F., we obtain the following rule : If the temperature exceed looo F., add to the actual density found by weighing *377 per lo F.of excess temperature over the loo0. It is obvious that the correction to be applied to fats, of slightly lower specific gravity, will scarcely differ from this. In practice it will generally be found that more accurate results can be obtained, by allowing for the error of temperature by this calculation, than by endeavouring to secure an absolute temperature of 1000 F.The diagram of the expansion of the fat will, however, be of assistance in enabling the specific gravity bubbles to be used for the purpose of taking the actual densities of fats. These bubbles expand under the influence of heat, very nearly in the same ratio as the ordinary specific gravity bottles ; in other words, notwithstanding the different conditions under which specific gravity bottles and bubbles are made, the cubical expansion corresponds very closely, and, consequently, in this respect no sensible error is incurred by the use of the bubbles, since the results obtained are practically identical with those which would be obtained by weighing in glass bottles.An accurate determination of the specific gravity of a bubble, at a temperature of 60°, d l be all that is necessary; or, if preferred, the actual specific gravity of the bubble may be f o l d by testing it on melted fat, the actual density of which has been accurately determined at looo, and calculating the result obtained from the expansion of the fat.I need hardly point out that it is useless to rely on any approach to accuracy in the bubbles as sold. I believe this inaccuracy is mainly due to the bubbles being marked before the glass has assumed its permanent form.If the sample of fat to be tested is gently and slowly heated, until the bubble of known specific gravity begins to fall, and the temperature of the fat noted, a glance at the diagram will shom’at once the agtual density of the fat at looo F. Every degree F. difference of temperature corresponds to -377 of actual density, and, as it is perfectly easy to determine the sinking point to *5 F., the actual density can in this way be ascertained to within *20, which is as near an approach to accuracy as is-requisite in most cases, or even practicable, from a single experimental weighing. Should the specific gravity determinations not coincide with those of a genuine fat, a further anrtlysis is in any case essential.