116 



CHEMISTRY. 



termined by Drs. Kriiss and Moraht by means 

 of what is probably the purest oxide ever pre- 

 pared. The value obtained, 9-05, is very nearly 

 a whole number, being nearer to 9 than any 

 value ever previously obtained. It therefore ap- 

 pears that the whole number 9 is more nearly 

 approached the purer the materials are with 

 which the experiment is performed a result 

 that has an important bearing in favor of Prout's 

 hypothesis. 



' The atomic weight of copper has been deter- 

 mined by T. W. Richards from the analysis of 

 cupric bromide, and found that of silver being 

 108 to be 63-644. 



Relations have been detected by M. Prud'- 

 homme between the shades obtained by the use 

 of mordants in dyeing and the atomic weights 

 of the substances. The shades appear to under- 

 go continuous variations, which appear very dis- 

 tinct by the side of Mendeleefs classification of 

 the elements in natural groups and periodic se- 

 ries. From numerous experiments made with 

 some thirty-six elements, the author concludes 

 that in each of Mendeleefs groups, if we con- 

 sider the terms of the periods of the even or odd 

 rank, there is a continuous variation in a deter- 

 mined direction from blue to red or from red to 

 blue. The author continues his paper with more 

 specific accounts of the variations. 



Chemical Analysis. Prof. Roberts-Austen 

 expressed an apprehension in his address at the 

 chemical section of the British Association that 

 the wide range of study upon which a metallur- 

 gical student is rightly expected to enter may 

 lead to diminution in the time devoted to ana- 

 lytical chemistry, and this most serious question, 

 he said, should be pressed upon the attention of 

 all who are responsible for the training of our 

 future chemists. There can be no question that 

 sufficient importance is not attached to the esti- 

 mation of " traces," an analysis being considered 

 satisfactory if the constituents found add up to 

 99-9. although a knowledge as to what elements 

 represent the missing 0-1 may be more useful in 

 affording an explanation of the defects in a ma- 

 terial than all the rest of the analysis. This 

 matter is of growing interest to practical men, 

 and may explain their marked preference for 

 chemists who have been trained in works to 

 those who have been educated in a college lab- 

 oratory. 



It has been found by Prof. E. Shulze, of Zu- 

 rich, and W. Maxwell, of Harvard College, that 

 in the estimation of fatty matters in vegetable 

 organisms the substituted glycerides do not be- 

 come wholly separated by extraction with ether 

 even in a great duration of time. If the mate- 

 rials which had been already extracted with 

 ether were' still further extracted with absolute 

 alcohol, another portion of substituted glycer- 

 ides was obtained, which in most instances was 

 greater than the amount separated by the ether. 

 The process of the alcohol extraction consisted 

 in merely extracting the material already treated 

 with ether and evaporating off the alcohol, and 

 re-extracting the alcohol extract residue with 

 ether. The reason for taking up tht s lecithins 

 out of the alcohol extract; was that it had been 

 observed that those lecithins which were orig- 

 inally insoluble in ether became soluble in that 

 menstruum when previously acted upon by alco- 



hol. In cotton seed the proportion of lecithin 

 to the total fatty bodies is very small, yet about 

 50 per cent, of those substituted glycerides were 

 left in the material after extraction with ether 

 for a period of fifteen hours. In certain other 

 varieties of seeds, such as the legumes, beans, 

 peas, vetches, etc., the total ether and alcohol 

 extracts were composed of lecithins varying 

 from 25 to 45 per cent. 



The method of J. Weirich for detecting col- 

 oring matters fraudulently added to wine is 

 founded on the action of aif and light upon the 

 coloring matters of wine spread out in an ex- 

 tremely thin layer. The wine is applied with a 

 brush upon a piece of paper of good quality 

 which is not pervious. It is kept for an instant 

 in contact with the paper, which is then drained 

 off and let dry. The coloring matters of wine 

 give the paper a different tone from those of 

 the coloring matters vegetable, animal, or arti- 

 tificial which serve for the sophistication of 

 wines. The natural colors of wine and vegeta- 

 ble colors are transformed upon paper, each ac- 

 cording to its nature. The artificial colors are 

 transformed either very slightly or not at all. 

 The process of A. Pagnoul depends on the prop- 

 erty of soap lyes to destroy the natural coloring 

 matter of wines without giving them the green- 

 ish tint communicated by other alkaline solu- 

 tions and without affecting strange colors. For- 

 eign coloring matters which are not decomposed 

 by alkalies at an ordinary temperature (like sa- 

 franin, eosin, etc.), or which they turn to a vio- 

 let (like the tropeolins and cochineal), are de- 

 tected by L. Sostigni's method by shaking the 

 wine for five minutes with a solution of potassa 

 one tenth its volume, and pouring the liquid 

 into a filter of parchment paper in contact ex- 

 ternally with water. After some hours a yellow 

 liquid containing the oxidation products of the 

 tannin is diffused. The coloring matters foreign 

 to wine are fixed upon the parchment paper 

 with their own colors. Natural wines color the 

 paper yellow. 



The reaction with nitric solution of ammo- 

 nium molybdate is applied by G. Deniges for 

 distinguishing arsenical spots from those of an- 

 timony. The suspected spots are mixed with a 

 few drops of nitric acid ; they dissolve instantly, 

 whether they consist of antimony or arsenic. 

 Heat is applied for a few moments and a few 

 drops of ammonium molybdate in a nitric solu- 

 tion are applied. When arsenic is present, even 

 if only in traces, a yellow precipitate soon ap- 

 pears, showing the forms of crystals fine yel- 

 low stars with triangular branches, generally six 

 in number, arranged in rectangular planes ac- 

 cording to the axes of a cube charactei-istic of 

 ammonium arsenio-molybdate. Antimony gives 

 nothing analogous. 



Gr. Vortmann shows that in the determination 

 of metals by electrolysis of their solutions it is 

 essential, first, that the metal be separated out 

 quantitatively, as such or in the state of a known 

 compound; and, second, that the precipitate de- 

 tained forms a uniform coating on the platinum 

 capsule used as an electrode and adheres so firmly 

 that no loss takes place on rinsing with water 

 and alcohol, and that it undergoes no change 

 during drying. Among the metals that have 

 been hitherto determined electrolytically, iron, 





