physics. After attending the classes a selection will be made of the most promising boys, who will then pass through a special course of technical lectures in glass, its chemistry and properties. After this they will be drafted into the works, where a special laboratory is being fitted up in which they, for a certain number of hours a day, will learn the rudiments of the many branches of the industry. The remaining portion of the day they will be employed in making themselves useful in one or other of the many departments of the firm. During this time of probation and instruction they will receive a small nominal wage, and as soon as, by their skill and industry, they are able to do useful work they will be passed into the works proper, and be paid the union wages for the duties they are fitted for. The speed at which they will learn their business will depend upon their own efforts aided by all the assistance that can be given to them in the way of instruction. experimental data are given at about that temperature and at ordinary atmospheric pressure. Since so much depends upon the relative volumes of atoms and molecules it is necessary to obtain these constants correctly. No one can assert that both Perkin's and Bruhl's results with regard to Allyl bromide, e.g., in Magnetic Rotation and Refractivity respectively are correct, for the former employed Specific gravity 142532 at 15° and the latter 141057 at 25° 13980 at 20° There is a difference of about 15 per cent, in the two cases, and therefore the same percentage difference in the relative volumes. In other cases there is frequently a difference in their specific gravities of about 1 per cent. Many scientists seem to think that Kopp solved the problem of relative volume, and that other people have slightly improved upon it; so it is necessary to go into the matter here, though as shortly as possible. Kopp started by finding the relative volume of CH, at o° C. to be 15.8. He then subtracted six times CH, from the total volume of hexane and found that the remaining two atoms of H together had a volume of about 30. He thought he had come to an absurdity, so he transferred his operations to the boiling points. Dr. Cohen in his "Organic Chemistry,' published in 1913, says: "The value of CH, can be determined by a comparison of different homologous series. The following table gives the constants for different classes of compounds at the boiling point and also the maximum and minimum values observed: Amines (aliph.) Alkyl iodides where n is the index of refraction. The molecular magnetic rotation (M) of a pure compound is given by the formula: where a is the magnetic rotation. It is evident that the last two sets of constants are dependent upon the first set, viz., the molecular volumes. It has been discovered that the relative volumes (when correctly found) of the elements reveal the constitution and structure of atoms in a definite manner. An example of this was given in the CHEMICAL NEWS, July 18, 1919. It has also been discovered that the constants obtained from the heat of combustion are dependent in a particular manner upon the relative volumes of the elements. Therefore, the Relative Volumes of the Elements are the Fundamental Constants of Nature. These constants are invariable only at a constant temperature, so that it is necessary to choose a definite temperature. The ordinary one of 15° C. is most convenient as the great majority of It It will be seen that there is not only a considerable variation in the constants in the different series, but a wide divergence in the minima and maxima for the same series. should, however, be pointed out that as a rule the value for CH, increases in ascending the series, and the longer the series and the wider the range of boiling points the larger the variation. There is clearly some missing factor which should be introduced to bring the various constants into uniformity." These, then, are the data upon which Kopp founded his "discovery" that-the relative volume of CH, is about 22 at the boiling points of all substances containing it. And yet Dr. Cohen says: "With a delicate balance it is possible to estimate specific gravities with accuracy to one part in 100,000, and, as this is the principal factor involved in the determination of molecular volumes, the method experimentally leaves little to be desired." He then sums up all the attempts at the solution of this problem in the following words: "As no constant values can be attached to carbon, hydrogen and oxygen, it seems useless to attempt to derive a molecular volume by the summation of atomic volumes." Dr. Cohen continues: "When Kopp's values for C, H, and O were calculated for compounds of the formula CaHbOc it was seen that the molecular volumes did not invariably conform to the calculated values and the property was therefore not strictly additive." Then he says: "Kopp showed subsequently that the calculated value would correspond with the observed result if account were taken of the oxygen atom The value for doubly linked oxygen was found a little higher that that for one atom of carbon On this new basis the calculated values for fortyfive different compounds did not vary by more than 4 per cent." Now suppose there are two substances, A and B, which are similar in all respects, except that A has 5 methylene groups in its constitution and B 6; and suppose A boils at 120° C. and B at 150° C. At 120° the volume of CH, in each substance will be the same, and as B rises to its boiling point, the volume of each CH, is increasing. Therefore the volumes of CH, in the two substances are different at their boiling points. Therefore it is impossible to compare the volumes of substances at their boiling points. Explanation of the Graph. The numbers 6-18 at the bottom of the graph represent the normal paraffins; e.g., 9 stands for C.H2, 12 for C12H2, and so on. The numbers Rel vol of CH, at 0° = 16 2 Rel vol of CH, at 1st. 16.42 Rek vol of CH, at 100° 1697 Forty-five out of several hundred, and correct within four per cent! If chemists are satisfied with this, of course it is no use discussing the matter further. Fortunately for Kopp, Maxwell and Clausius discovered that-at their respective boiling points and at atmospheric pressure the molecules of various liquids occupy a space very nearly 03 times the total apparent volume. So they decided that he was right in taking the various boiling points as the stepping-stones from which to make his discovery. So it is necessary here to produce a proof that it is impossible to compare the relative volumes of substances at their boiling points. Kopp found that at o° C. the volume of CH, in all substances is 15.8. The correct value is 16.2. At 15° C. it will be shown later that the volume of CH, in hundreds of substances is always 16'42. At 100° C. it is 16'97 (see graph). And at any other arbitrary temperature the volume is constant, gradually increasing with the temperature. If another point be taken on the same curve, say, where the specific gravity of CH is 0.780, the relative volume is 253.85. 30 Subtract 140 45 from 253.85 and divide the result by 7 because there is a difference of 7 times CH, in the two molecules, the quotient is 16:2, which is the relative volume of the group CH, at o°. Similar results will be obtained by taking any two points on the other two curves. Now it is quite plain from these three curves that the graphs at any other arbitrary temperatures, say, 50°, 150°, and 200° would be similar curves drawn in similar directions, and moreover 15° 0.6825 0678 155 Schorlr 0.6985 14° Lachowicz 0'7022 07023 14° Lachowicz 0'7177 0.719 17°.5 Schorlr 07165 15° 6 Thorpe 0'7304 0'731 16° Lemoine 07330 0° Le Bas 0.7408 07324 20° Lachowicz O'753 16° Lemoine 0'7497 07448 15° Krafft 07581 20° V.N. 0'7572 07548 15° Krafft 07684 20° V.N. 0.7637 0.7608 15° Krafft 0'7694 07681 15° Krafft 0'7744 07724 15° Krafft 0.7788 07754 18° Krafft 07707 25° Krafft 0.7828 07764 22°5 Krafft 07714 30° Krafft 0.7863 07768 28° Krafft 0.7895 07288 99° Krafft O'732 curves of this nature are the only ones which would give constant volumes for CH2. But the graph of the specific gravities at the boiling points (as shown) does not lie anywhere near the general direction of these curves, and therefore it is quite evident that the relative volume of CH, at the boiling-points is not anything like constant. The table contains the experimental data, which are required for the graph. There being no regularity in the relative volumes of substances at their various boiling points, if absolute regularity can be demonstrated to exist at a fixed temperature, of course the question as to which method of procedure is correct will be settled. And if this regularity produces an independent proof of the Constitution and Structure of the Chemical Elements as given in previous papers, the matter will be doubly settled; for the relative volumes, as supposed to have been obtained from the various boiling points, lead to absolutely nothing. Again, if this regularity leads up to the demonstration of the Correlation of Physical Constants, such as those mentioned at the commencement of this paper, of course the matter will be trebly settled. By the action of ethy! nitrate on phenyl magnesium bromide a deep blue solution was obtained which has the properties of an indicator, turning pink with acids and blue with alkalies. One equivalent of ethyl nitrate C,H,ONO, was added gradually to one equivalent of C,H,MgBr, cooled in ice. In a few minutes a vigorous reaction takes place, with the evolution of heat. After standing over night, pieces of ice were added and the mixture left for a few hours for the magnesium compounds to settle. The etheral layer was then separated and the ether evaporated off. Solid Na,CO, and water were then added, January 7, 1921 when the mixture turned blue (later work shows that the colour is sometimes not developed till later). The mixture was then extracted with chloroform, when it was found that the blue colour remained in the aqueous layer. This blue solution, which is obviously the sodium salt of the indicator, cannot be kept very long except as a very dilute solution. It was found that the best way to keep the substance was as an ethereal solution of the free indicator, so the blue solution was acidified, and the free red indicator extracted with ether. The presence of chloroform seems necessary for the formation of the blue salt; all attempts to obtain the substance by extraction with ether resulting in compounds of the nitrolic acid type. 215 cc. NaOH solution neutralised 25 cc. N/10 H2SO.. The end point with the new indicator is very sharp and easy to observe, the last drop of acid or alkali making a radical change in colour. Three titrations did not differ from each other by more than o'05 cc. The substance is, unfortunately, not very stable, losing its colour and indicating properties on keeping. On concentrating a water solution of either the free indicator or the sodium salt, the substance also loses its colour and properties. On allowing an ethereal solution of the free indicator to evaporate, a few drops of a dark red liquid were obtained, which soon decomposed. Both the free indicator and the sodium salt are very easily reduced by SO,, but H2O2 seems to have no effect. The depth of the colour indicates that the substance is of high molecular weight. Evidently there is some fundamental change during the reaction, as the ordinary course of a Grignard reaction will not give a final substance which shows the necessary tautomerism to account for the indicating properties, i.e., from theory we would expect : IN certain respects bracken rhizomes resemble potatoes in composition. In Table I. some average figures for potatoes are quoted from the recently issued Report of the Food (War) Committee, Royal Society (Report on the Composition of Potatoes grown in the United Kingdom. Food (War) Committee, Royal Society, 1919). The figures for "Arran Chief" are given as those of a potato of good quality, above the average in dry matter and in nitrogen. As compared with the potato-tuber, the bracken rhizome contains more dry matter, but the dry matter is not quite so high in soluble carbohydrates, and is much fibrous. In albuminoids it has already been seen that bracken rhizomes vary greatly. Potatoes are not so high in this valuable constituent as the B samples from Craibstone, but are higher than the A samples and than the samples in Table I. more In the case of the potato the main part of the solid matter consists of starch, which forms most of what is called "soluble carbohydrates." Bracken rhizomes also contain starch in considerable quantity, but, unlike the potato, the larger part of the soluble carbohydrates does not consist of starch, but of other substances of an undetermined nature. In order to test whether animals would eat bracken rhizomes, and if so whether they would thrive upon them, a number of feeding trials were made with different classes of stock. It had already been reported from various sources that pigs will root up and eat bracken rhizomes, and it is recorded that these have been used in Europe as human food in times of scarcity. It is also on record that they formed a portion of the food of the natives of New Zealand at the time of the discovery of that country by Captain Cook, Attempts were made in our experiments to feed these rhizomes to pigs, to cattle, and to poultry. On account of the many difficulties of the times it was almost impossible to arrange accurate feeding trials in 1918. Many people were suffering from lack of food stuffs, and were quite ready to give any promising new substance a trial, but very few had the necessary time and labour for weighing out foods and weighing stock, and otherwise giving the attention required for carrying out a reliable feeding experiment. Some rough trials were first made, and these showed that at any rate bracken rhizomes were not very palatable to stock. Pigs would not eat them if they were already accustomed to a liberal diet, especially of concentrated foods. They took them only when they were in lean condition and, generally speaking, breeding stock consumed them more readily than feeding stock. It was also discovered that pigs ate them more readily if the rhizomes were given to them in the rough state, just as they were dug from the soil. It was at first assumed that if they were cleaned, chopped, mixed with other food, and cooked, the animals would consume them more readily. This proved to be quite wrong. Whether mixed with other The rhizomes which were dried were dug in May, when considerable growth of frond had already taken place, and they were not therefore of such good composition as if they had been dug a month earlier. It was found impossible, in view of the labour conditions then prevailing, to have them washed. This accounts for the high percentage of ash and sand which they contained. Some feeding experiments with pigs and poultry were carried out with the help of members of the staff of the West of Scotland College of Agriculture, and the results are detailed in Bulletin No. So of the College, "Report on Experiments of the Feeding Value of Bracken Rhizomes for Pigs and Poultry." Preliminary trials were made at Holmes Farm, Kilmarnock, in which it was shown that breeding sows would eat considerable quantities of the rhizomes without any evil effect, but that well-fed feeding animals could not readily be induced to eat them. Then, through the kindness of the Directors of the Scottish Co-operative Wholesale Society, and of Mr. Geo. C. Young, their manager, a more accurate test was made on two pens of pigs at Calderwood Estate, East Kilbride, under the supervision of Mr. James Wyllie, B.Sc., of the West of Scotland College. These pigs had been somewhat short of food, and as their appetites were not pampered, they ate their allowance of bracken, whether in the form of untreated rhizomes or ground into meal, with apparent relish. But the results of the experiment were unsatisfactory, as the pen of pigs which got bracken in addition to the foods given to the other pen showed very little benefit from the extra food. The pigs which got bracken made a live-weight increase of o77lb. per head per day over a period of 140 days, while those without it made an increase of o71lb. per head per day. The small increases made all over were due to the poor diet on which the pigs were kept owing to the scarcity of feeding-stuffs, but the pen which had from 1lb. up to 2lb. of bracken meal per head per day as an addition to their ordinary diet did very little better than the others. Mr. James Picken, Milton, Kircudbright, also tried both the untreated bracken and the meal with pigs, but was not in a position to do any weighing. He reported, however, that the pigs ate the bracken "with evident relish," and that it could "well replace other fresh vegetable food Mr. during the winter and spring months." Picken also noted that the bracken appeared to make the pigs lively, and that those which received it took more exercise than the others. He remarks: "This is a decided advantage to pigs 6 weeks to 4 months old, but over that age and during the fattening period it might be necessary to withold the roots, as these might unnecessarily excite the pigs and thus delay the fattening process." The evidence, so far as it goes, indicates that bracken rhizomes in the fresh state or as meal might be of some use as a maintenance food for pigs, but are not of much use as a fattening food. With poultry the results were still less favourable. The experiments were carried out with bracken meal at the Poultry Department, Holmes Farm, Kilmarnock, by Miss Agnes Kinross of the The West of Scotland Agricultural College. birds did not like the bracken meal even when it was mixed with a considerable amount of other food, and the pen which received a little bracken in their food during October and November did not lay so well as others which had received no bracken meal but an equivalent in other foods. The experiment indicates that bracken meal is not a suitable food for egg-production. Bracken meal was tried with feeding cattle at Craibstone, the experimental farm of the North of Scotland Agricultural College of Agricutlure. The quantity available was not sufficient for an experiment on a scale large enough to be satisfactory, as for that purpose 8 or 10 bullocks would require to have been fed with the meal for a period of several weeks, and the quantity should gradually have been raised to several pounds per head per day. The quantity available was only sufficient to enable two bullocks to get 1lb. per head per day for two months. A comparison lot of two bullocks was similarly fed, but without bracken meal. Lot (bracken meal) received daily 95lb. turnips with oat straw ad lib., and 2lb. bruised oats, 2lb. cereal offals, and 1lb. bracken meal. Lot 2 received daily the same amount of turnips and straw and 24lb. each of bruised oats and cereal offals. Lot 2 therefore received lb. oats and lb. cereal offals instead of 1lb. bracken meal given to Lot 1. Both animals in Lot I ate the bracken meal with evident relish, and they would have taken more had we been able to supply it. The increases made by the two lots were somewhat similar. In 56 days the cattle in Lot I made an average increase of 132lb., or 2.4lb. per head per day, while those in Lot 2 made an average increase of 113lb., or 2.0lb. per head per day. The lot which got the bracken meal therefore did slightly better than the other, but no weight can be attached to this result, as the number of animals under experiment was too small, and the amount of bracken meal fed was only a small proportion of the whole food. So far as it goes, however, the experiment shows |