1. The petitioner has been convicted by the Second Presidency Magistrate of selling adulterated butter and thereby committing an offence punishable under Section 5, Clause (1)(a) and (b) of the Madras Prevention of Adulteration Act, read with Rules 28 and 29 under the same Act. He has been sentenced to pay a fine of Rs. 100.

2. The prosecution case rests entirely on the evidence of the Corporation Analyst, with the figures that he gave for certain tests carried out by him and more important still, the inferences which he drew from the figures obtained. The accused examined on his side a gentleman who has had some experience of analytical work under the Bombay Government; and fortunately the figures that he obtained upon analysis were very much the same as those given by the Government Analyst in so far as the tests conducted by these two gentlemen were the same. The Government Analyst conducted four tests. He estimated the Beichert Wollny value, the Befractive Index, and the Saponification value; and he then calculated from the Beichert Wollny number and the Saponification value the Juckenack and Pasternack value. D.W. 1 estimated the Reichert Wollny value and the Saponification value and the Refractive Index. He did not calculate the Juckenack and Pasternack value; but he carried out three other tests to estimate the Polensky value, the Iodine value, and the Phytosteryl Acetate number. The Reichert Wollny value for the specimen in question was found to be 21.8 by the Corporation Analyst and 21.91 by D.W. 1. The Refractive Index was found to be 1.4549 and 1.4545 by these two gentlemen respectively. The Saponification values found by them were 216.9 and 217.8 respectively.

3. The Corporation Analyst deposed that the Reichert Wollny number varies in Madras from 28 to 40 for buffalo's milk and from 25 to 35 for cow's milk; so that the value obtained for this particular specimen was considerably below the minimum which he considered to be consistent with its being pure cow or buffalo's fat. The Refractive Index, he says, varies from 1.4525 and 1.4565. The Refractive Index of the specimen in question was found to be 1.4549, which lies about half way between these two extremes. This in itself is merely a negative result; for the Refractive Index can be easily manipulated by balancing the kinds of fats that are used in adulteration. The Saponification value was found to be 216.9, whereas the lower limit which the Corporation Analyst considers to be consistent with the purity of the specimen, is 219. It is thus seen that both as regards the Reichert Wollny value and the Saponification value the specimen gave results which the Corporation Analyst considered could not be obtained from pure butter, whether obtained from cow's milk or buffalo's milk. He estimated that the degree of adulteration was 20 per cent., which percentage he obtained by subtracting the Reichert Wollny number from the minimum permissible, viz., 28, and calculated the difference as a percentage of 28. The purport of D.W. 1's evidence is that in practice the Bombay Government did not prosecute unless the Reichert Wollny number fell below 19 and that as this sample was not below 19 and the Saponification value was not unduly low, especially having regard to the lownesss of the Reichert Wollny number, with which it increases or decreases, it was unsafe to assume that the specimen was not pure butter.

4. There is no doubt in my mind that the figures found for this specimen of butter were abnormally low for pure butter. Reference was made here in the argument to the discussion of these tests in a number of text-books, such as, Warburton's "Chemical Technology," Analysis of Oils, Fats, and Waxes by Lewkowitsch, Bolton's 'Oils, Fats and Fatty Foods,' Woodman's 'Food Analysis', Allen's "Commercial Organic Analysis," and Thorpe's "Dictionary of Applied Chemistry." All these treatises indicate that the normal Reichert Wollny value and Saponification value are what are considered reasonable for a pure specimen by the 'Corporation Analyst. Allen states that in some tests that were made of ghee made from butter produced from the mixed milk of buffaloes collected from various dairies in different parts of India, the minimum Reichert Wollny value found in any specimen was 29, and the lowest Saponification value 226, as against the figures of 21.8 and 216.8 for the specimen in question. Of 357 samples examined by Thorpe, the average Reichert Wollny number was about 28.5 and the average Saponification value about 228. Bolton, at page 117, says that the Saponification value usually falls between the limits of 226 and 232. Wherever lower numbers were found, the results are referred to in the text-books as abnormal.

5. The prosecution rely on a test which does not find much mention in the text-books, and which is known as the Juckenack and Pasternaek test. The Juckenack and Pasternack number is obtained by subtracting 200 from the Saponification number and deducting the figure so obtained from the Reichert Wollny number. It is stated in Thorpe and Woodman that the limits for that number lie between + 4.25 and -3.50, whereas the figure obtained for the present specimen was slightly higher than the upper limit, namely, 4.9. There is however little discussion of this test even in these two books and none at all in the others. I would therefore prefer not to express any opinion on the significance of the number (4.9) found for this specimen. It may however be noted that some of the specimens of butter tested by Thorpe had a Juckenack and Pasternack number as high as + 5.6.

6. Although the Reichert Wollny and Saponification values of the specimen are abnormal for pure butter, the text-books make frequent references to instances in which very much lower numbers were found; and, most books emphasise the very wide limits between which the Reichert Wollny and Saponification numbers vary. In the table given by Thorpe for his examination of the 357 samples above referred to, seven specimens had numbers as low as 22.5 and 219.9 respectively, while ten were as high as 32.6 and 232.5. Although it is said at p. 814 of Lewkowitsch that in the case of ghee prepared from the milk of Indian buffaloes, T.K. Ghose found Reichert Wollny values ranging from 29.2 to 42 with an average of 34.5, and that it was proposed that 28 should be regarded as the reasonable minimum Reichert Wollny value, yet some specimens examined by Browning and Parthesarathy gave figures as low as 21. In a table at p. 835 of Lewkowitsch are found values as low as 17. The learned Counsel for the appellant relies above all on figures found in a paper contributed to a series of Government publications entitled "Memoirs of the Department of Agriculture in India". The title of the contribution is: "variations in some characteristics of the fat of buffalo and cow milk with changes in season and feeding" by F. J. Plymen and A. R. Padmanabha Aiyar. Connected with it is an article entitled "The mutual applicability of the analytical figures for butter, fat and ghee". Various animals or groups of animals were examined by the contributors, and a herd of buffaloes, described as the Telenkheri herd, were found to give very low Reichert Wollny and Saponification numbers, the former falling as low as 18.72 and the latter as low as 212.1. Even the highest number in this series of examinations does not approach the minimum standard which the Corporation. Analyst considers consistent with the purity of the sample. The fact that the defence relied on a contribution as old as this one which was in 1921, is however an indication, I think, of the abnormality of the figures for this Telenkheri herd. Many such researches must have been made since 1921; but apparently no later figures have given the support to the accused that these particular figures give. We do not know whether the writers of this paper are authorities on the subject; and I should be rather reluctant to attach undue importance to their figures, were they not rendered probable by the fact that very low figures are not infrequently obtained in all parts of the world. So while it is certain that the figures in the sample are abnormal for pure butter, it is difficult to be sure beyond all reasonable doubt that these figures are quite inconsistent-even in Madras-with the purity of the sample.

7. It has been pointed out that it is almost certain that the butter from which the sample was taken was a mixture of butters from various cows and herds and probably not all from the same villages, and that it is therefore most unlikely that abnormally low figures would be found for such a mixture. It is true that if butter with abnormally low Saponification and Eeichert Wollny numbers is mixed with a large number of other butters, the average for the whole mixed butter is likely to depart only very slightly from normal; but we have not the material before us to know the exact extent to which the butter from which the sample was taken is a mixture of various butters. We do not know from where the butter came, or the characteristics of the cows or buffaloes or herds from which the milk was taken. It is also possible, though perhaps not very probable, that the accused was unfortunate in that the samples were taken from a part of the bulk which represented butter of a cow or herd bearing abnormally low Values.

8. The prosecution rely on a case reported in Grande Venkata, Ratnam v. Corporation of Calcutta (1920) I.L.R. 47 Cal. 633 in which the prosecution had proved that certain samples of ghee were adulterated. The figures for the samples in that case were: Reichert Wollny Nos. 26.6 and 24.9; Butyro-Refractive Index Nos. 44 and 45; Saponification Nos. 222 and 220. Further, a colour test for adulteration with a vegetable oil gave positive results in three out of four trials. The prosecution point out that in that case, where the learned Judges accepted the contentions of the Corporation Analyst, the Eeichert Wollny and Saponifieation numbers were appreciably higher than those in this case; but two other tests were applied in that case which also indicated that the ghee was adulterated. The Refractive Index was there abnormal, whereas in the present case it is normal. Further, a delicate colour test was applied which indicated positively the presence of vegetable oil, not a natural constituent of butter. Any abnormality of the Reichert Wollny and Saponification numbers is sufficient to arouse an analyst's suspicions and to lead him to> carry out other tests. If they also indicate that the ghee was adulterated, then the result of all the tests might put the matter beyond all possible doubt. Unfortunately, an abnormal Reichert Wollny number is in pure butter associated with an abnormally low Saponification number; for the Reichert Wollny number depends on the quantity of Volatile fatty acids, such as Butyric Acid, while the Saponification number depends upon the total quantity of volatile and non-volatile fatty acids; and abnormality does not necessarily mean adulteration.

9. Both the prosecution and the defence have stressed the necessity for the laying down of a reasonable standard by the Government, such as has been laid down in other provinces. The reason why the Government of Madras has not done so is doubtless the difficulty in fixing a reasonable number, as the above analysis has shown. If too low a number is fixed, then it tends to give legislative approval to a high degree of adulteration. On the other hand if the number is high, a few pure butters and ghees could not be sold. It is however clear from the various text books that the Reichert Wollny number is considerably influenced by the conditions under which the animals live; and in most cases the Beichert Wollny number can be raised to any reasonable standard that may be laid down by the Government by modifying somewhat the living conditions of the animals and by not using for the preparation of buttermilk secreted towards the end of lactation. Even if keepers of herds did not do this, very little hardship would be caused by the Government's fixing a number in the neighbourhood of 30; for very few genuine butters would be excluded. On the other hand, the danger of having such a low number as 19, which has been adopted in some provinces, would be that a vendor of butter could very appreciably adulterate his butter without Very much risk to himself; for Courts would be inclined to feel, in the absence of strong evidence by the prosecution to the contrary, that if that particular standard was reached, the accused would be entitled to an acquittal. In the Calcutta case above referred to, the Court suggested that the figure of 30 for the Reichert Wollny number was reasonable; and it appears that that number was adopted soon afterwards by the Government of Bengal, although they subsequently reduced it from 30 to 28. 28 is the figure which the Corporation of Madras has for a considerable time regarded as the minimum consistent with the purity of the butter examined, and it would seem inadvisable to fix the number below that figure. In India, most butters are made from buffalo milk, which has appreciably higher numbers (10 to 12 per cent, higher at least) than cow's butter.

10. Although, therefore, it is very probable that the accused did adulterate his butter with an appreciable percentage of foreign fat and oil, yet I do not consider that it has been proved beyond all reasonable doubt that the accused was guilty of the offence with which he was charged.

11. The appeal is allowed and the fine ordered to be refunded.