neglecting to apply the corrections thus ascertained. He further concludes that polariscopes equipped with a Reichsanstalt scale should be controlled by the incorrect Reichsanstalt standard in order that the plates shall agree with the instrument scales. In answer to this, we submit the obvious fact that the control plate should be used to standardize the scale and not the reverse.

The associate referee further remarks that most sugar chemists who have gone into the subject are free to admit that the 100° point is low, but there is grave doubt whether it is as low as the Bureau of Standards makes it. The latter part of this assertion is evidently based upon his further statement that "The work upon which this value was founded has been severely criticised by sugar chemists abroad". The writers wish to enter a solemn protest against any such misstatement of facts. So far as our information goes, there has only been one statement which could be construed as a criticism of the work of the Bureau of Standards in determining the 100° point of the saccharimeter. The criticism referred to is a statement by A. Herzfeld1 (Institut für Zuckerindustrie, Amrumstrasse, Berlin, N. 65, Germany), who was the joint collaborator of Otto Schönrock (Physikalisch-Technische Reichsanstalt, Berlin, N. W., Germany) in determining the previously accepted value of the 100° point. The so-called criticism is approximately 150 words in length and is confined to a short discussion of several minor points involved in the preparation of the sugar used by the Bureau of Standards. Herzfeld merely conveys the idea of a possibility of the existence of a slight error due to these causes. He does not furnish the slightest proof that any error was made. On the contrary, all measurements and all contributory data developed since the determination of the constant by the Bureau of Standards have merely served to verify the correctness of the new value. Fortunately, or unfortunately, depending upon the point of view taken, the correction of the error in the old value of the 100° point operates in favor of the producers of sugar and against the buyers of sugar. For many years the buyers of raw sugars have had the advantage of a lowering of the test of over 0.01S.

In view of these facts, the writers desire to advise against the adoption by this association of the associate referee's recommendation that members sending quartz plates to the Bureau of Standards for standardization request that the Bureau of Standards certify on the old value of the 100° point in place of the present corrected value until this has been adopted internationally.

1 Z. Ver. Zucker ind., 1917, 67: 407.

THE ATTITUDE OF THE NEW YORK SUGAR TRADE UPON THE NEW BUREAU OF STANDARDS VALUE FOR

STANDARDIZING SACCHARIMETERS1.

By C. A. BROWNE (New York Sugar Trade Laboratory, New York, N. Y.).

The question of the standardization value of normal quartz plates for saccharimeters, which A. H. Bryan discusses in his report, page 326, is a matter of extreme importance, as the difference of 0.1 sugar degree between the official standards of Berlin and Washington will mean a difference of over one million dollars in the valuation of sugars handled by the New York Sugar Trade. This conflict of opposing standards has already caused some uneasiness in the trade, and the matter has been carefully considered by the Directors of the New York Sugar Trade Laboratory, who represent equally the buyers and sellers of raw sugar.

The standard of graduation for the German or Ventzke scale which has been employed by the New York Sugar Trade up to the present time is the one which is being used by practically all members of the International Commission for Uniform Methods of Sugar Analysis, the methods of which are the official methods of the New York Sugar Trade. The Directors of the New York Sugar Trade Laboratory, after careful consideration of the question, voted unanimously against the adoption of the new Bureau of Standards value for standardizing saccharimeters for the following reasons:

(1) That the new Bureau of Standards value has been criticised by European investigators and until it has been confirmed by testing bureaus in other countries and has been agreed to internationally it would be exceedingly unwise to adopt a value which might have to be changed again to something else. The Directors of the New York Sugar Trade Laboratory do not wish to make any departure in methods or standards unless the proposed changes have a fair prospect of permanency, especially in international transactions, otherwise the sugar trade will be subjected continually to disturbances of this kind.

(2) Granting that there may be a slight error in the present German standard, and even granting that the new Bureau of Standards value may be correct, the Directors of the New York Sugar Trade Laboratory are of the opinion that no injustice is being done at present to the sellers of raw sugar for the reason that the minus error due to scale graduation is offset by an equal or greater error due to the volume of the lead precipitate in clarification.

1 Presented by G. L. Bidwell.

If a scale error exists it should by all means be corrected, and the true standardization value fixed by international agreement. When this scale error is corrected, the counter-balancing lead precipitate error should also be corrected either by dry lead defecation, as proposed by W. D. Horne, or by other accurate means.

I might say in conclusion that this decision of the Directors of the New York Sugar Trade Laboratory is in complete agreement with the opinion of Prinsen Geerligs of Holland and other European authorities.

COMMENTS ON RECOMMENDATIONS PROPOSED BY A. H. BRYAN.

By W. D. HORNE (National Sugar Refining Company, Yonkers, N. Y.). The following comments are presented relative to recommendations 3, 4, 5, and 6 of the associate referee on sugar, page 329:

(3) In mixing sugar samples, particular care should be taken to avoid excessive exposure to the air. Speed is of first importance. Avoidance of exposure of thin layers of sugar to the air may be helped by mixing in a mortar instead of on a plate.

(4) In cases where the invert sugar is known, more accurate results for individual samples may be obtained by using the formula P200 = Pt+ 0.0003°S (t −20) — 0.00812°L (t — 20) where S sucrose and L = levulose (one-half the invert sugar).

(5) The Baumé scale is so thoroughly entrenched in the industries that it will continue to be used and must be taken into consideration. The Bureau of Standards new scale and table is of particular value, having the modulus 145, which is close to the averages heretofore in use, is a round number and has already been accepted by the Manufacturing Chemists Association. The reference of sugar solutions at 20°C. to water at 20°C. is also a great advantage.

(6) The Bureau of Standards work on the 100° point of the polariscope was conducted with the utmost precaution and the one criticism of it is merely a suggestion that some inversion of the pure sucrose may have occurred. No evidence of such inversion is given, however, and it is not at all probable that it occurs, since the sugar solutions used never were heated above 35°C.

The government will doubtless continue to employ the corrective value and it is important that all remaining errors of polarization, especially elevation due to the volume of the lead precipitate, be eliminated as soon as possible.

A SUGGESTED MODIFICATION OF THE METHOD FOR CRUDE FIBER1.

By L. D. HAIGH (University of Missouri, Columbia, Mo.), Associate Referee on Crude Fiber in Foods and Feeding Stuffs.

The study of the crude fiber determination this year has centered around the filtration operations. The treatment of feeds with dilute acid and alkali produces solutions which do not pass easily through ordinary filtering media. This has led to the use of muslin or linen as filtering media. The uncertainty that the cloth holds all the fine particles of fiber and the fact that no two chemists would be likely to use the same variety of cloth in filtering, renders this mode of filtering undesirable. Because of these reasons, the writer has preferred the one filtration method since it avoids one of these doubtful operations.

The associate referee has been unable to find a quick filtering medium which could be substituted for the muslin or linen cloth. The feeds most difficult to filter in the crude fiber determination are those of high protein content, such as linseed and cottonseed meal. If the protein in the feed could be dissolved without producing a colloidal solution and removed before the acid treatment, the filtration could be conducted without difficulty, using asbestos, thereby avoiding loss of fine material. To test this, experiments were conducted with various proteolytic enzyms. The best results were obtained with pepsin. It is, therefore, suggested that a preliminary treatment with pepsin be made in order to render the protein soluble. It may then be separated from the undissolved material by filtration through asbestos. Subsequent treatment of the residue with 1.25 per cent acid and alkali will not produce solutions which are difficult to filter, such as are obtained with the old procedure. The cloth filtration can then be abandoned and a filtering medium used, such as asbestos, which is more certain to retain all the insoluble material.

The time lost in trying to filter the colloidal solutions from the regular method is largely saved, even though another step, which adds about 40 minutes' work to the process, is introduced. The results should be entirely comparable with those obtained by the old method and have the additional advantage of greater accuracy.

INSTRUCTIONS TO COLLABORATORS.

Two samples were prepared-silage and cottonseed meal. Both samples had been washed with ether and brought to a stable, air-dry condition before they were distributed. The collaborators were asked to run

1 Presented by C. R. Moulton.

each of these samples as follows: (1) By the regular official method; (2) by the regular official method, preceded by a preliminary digestion at 40°C. for 30 minutes, with pepsin in the presence of dilute hydrochloric acid, after which the residue should be filtered, using an asbestos filter; (3) by the one filtration method, preceded by the above-mentioned digestion with pepsin; (4) in addition to the above, the acid filtrate from Method 2 was carried through the one filtration method to determine how much material precipitated from this acid solution by the alkali digestion remains to contribute to the result on crude fiber.

Some wide variations are observed in the results of the nine collaborators with the method using pepsin. The results, especially with the cottonseed meal, indicate to some extent a lack of uniformity in the samples. A uniform mixture of the ingredients of cottonseed meal is difficult to make and maintain. However, every reasonable effort was made to avoid this error.

A further experiment conducted in the writer's laboratory by W. S. Ritchie and T. H. Hopper is shown in the table as Method 5. It was thought possible that the hydrochloric acid washing of the residue could be omitted in the one filtration method when the preliminary digestion with pepsin was made.

Two of the collaborators reported that the solutions were difficult to filter where pepsin had been used. This variation may be due to some collaborators having used pepsin, which was weak or utterly useless.

The writer believes that in the removal of the protein from the feed before the digestion with acid, can be found the solution of the errors of the crude fiber method.

RECOMMENDATION.

It is recommended

That Method 2 receive further study and that suggestions and criticism be solicited.