Constituents containing Silicon,' which contain the results of experimental investigations of equisetum species—distinguished for their exceptionally high 'ash' with large proportion of silica—to determine whether there are any grounds for assuming the existence of silicon-organic compounds in the plant, the analogues of carbon compounds. The conclusions arrived at are entirely negative. In reference to the second assumption that the cuticular tissues of cereal straws, of esparto, of the bamboo, owe their special properties to siliceous components, it has been shown by direct experiment upon the former that their rigidity and resistance to water are in no way affected by cultivation in a silica-free medium. In other words, the structural peculiarities of the gramineæ in these respects are due to the physical characteristics chiefly of the (lignified) cells of the hypodermal tissue, and to the composition and arrangement of the cells of the cuticle.

'Swedish' filter papers of modern make are so far freed from inorganic constituents that the weight of the ash may be neglected in nearly all quantitative experiments [Fresenius, Ztschr. Anal Chem. 1883, 241]. It represents usually about 1/1000 mgr. per 1 sq. cm. of area of the paper.

The form of an 'ash' derived from a fibrous structure, is that of the 'organic' original, more or less, according to its proportion and composition. The proportion of 'natural ash' is seldom large enough, nor are the components of such character as to give a coherent ash, but if in the case of a fibrous structure it is combined or intimately mixed with inorganic compounds deposited within the fibres from solution, the latter may be made to yield a perfect skeleton of the fibre after burning off the organic matter. It is by such means that the mantles used in the Welsbach system of incandescent lighting are prepared. A purified cotton fabric—or yarn—is treated with a concentrated solution of the mixed nitrates of thorium and cerium, and, after drying, the cellulose is burned away. A perfect and coherent skeleton of the fabric is obtained, composed of the mixed oxides. Such mantles have fulfilled the requirements of the industry up to the present time, but later experiments forecast a notable improvement. It has been found that artificial cellulose fibres can be spun with solutions containing considerable proportions of soluble compounds of these oxides. Such fibres, when knitted into mantles and ignited, yield an inorganic skeleton of the oxides of homogeneous structure and smooth contour. De Mare in 1894, and Knofler in 1895, patented methods of preparing such cellulose threads containing the salts of thorium and cerium, by spinning a collodion containing the latter in solution. When finally ignited, after being brought into the suitable mantle form, there results a structure which proves vastly more durable than the original Welsbach mantle. The cause of the superiority is thus set forth by V. H. Lewes in a recent publication (J. Soc. of Arts, 1900, p. 858): 'The alteration in physical structure has a most extraordinary effect upon the light-giving life of the mantle, and also on its strength, as after burning for a few hundred hours the constant bombardment of the mantle by dust particles drawn up by the rush of air in the chimney causes the formation of silicates on the surface of the mantle owing to silica being present in the air, and this seems to affect the Welsbach structure far more than it does the "Clamond" type, with the result that when burned continuously the Welsbach mantle falls to so low a pitch of light emissivity after 500 to 600 hours, as to be a mere shadow of its former self, giving not more than one-third of its original light, whilst the Knofler mantle keeps up its light-emitting power to a much greater extent, and the Lehner fabric is the most remarkable of all. Two Lehner mantles which have now been burning continuously in my laboratory for over 3,000 hours give at this moment a brighter light emissivity than most of the Welsbachs do in their prime.' ...'The new developments of the Clamond process form as important a step in the history of incandescent gas lighting as the discoveries which gave rise to the original mantles.'

It has further been found that the oxides themselves can be dissolved in the cellulose alkaline sulphocarbonate (viscose) solution, and artificial threads have been spun containing from 25 to 30 p.ct. of the oxides in homogeneous admixture with the cellulose. This method has obvious advantages over the collodion method both in regard to the molecular relationship of the oxides to the cellulose and to cheapness of production.

UNTERSUCHUNGEN ÜBER VERSCHIEDENE BESTIMMUNGSMETHODEN DER CELLULOSE.

H. Suringar and B. Tollens (Ztschr. angew. Chem. 1896, No. 23).

INVESTIGATION OF METHODS OF DETERMINING CELLULOSE.

Introduction.—This is an exhaustive bibliography of the subject, describing also the various methods of cellulose estimation, noted in historical sequence. First, the Weende 'crude fibre' method (Henneberg) with modifications of Wattenberg, Holdefleiss, and others is dealt with. The product of this treatment, viz. 'crude fibre' is a mixture, containing furfuroids and lignone compounds. Next follows a group of processes which aim at producing a 'pure cellulose' by eliminating lignone constituents, for which the merely hydrolytic treatments of the Weende method are ineffectual. The method of F. Schulze—prolonged digestion with dilute nitric acid, with addition of chlorate—has been largely employed, though the composition of the product is more or less divergent from a 'pure cellulose.'

Dilute nitric acid at 60-80° (Cross and Bevan) and a dilute mixture of nitric and sulphuric acids (Lifschutz) have been employed for isolating cellulose from the lignocelluloses. Hoffmeister modifies the method of Schulze by substituting hydrochloric acid for the nitric acid. Treatment with the halogens associated with alkaline processes of hydrolysis is the basis of the methods of Hugo Muller (bromine water) and Cross and Bevan (chlorine gas). Lastly, the authors notice the methods based upon the action of the alkaline hydrates at high temperatures (180°) in presence of water (Lange), or of glycerin (Gabriel). The process of heating to 210° with glycerin only (Hönig) yields a very impure and ill-defined product.

For comparative investigation of these processes certain celluloses and cellulosic materials were prepared as follows:

(a) 'Rag' cellulose.—A chemical filter paper, containing only cotton and linen celluloses, was further purified by boiling with dilute acid and dilute alkali. After thorough washing it was air-dried.

(b) Wood cellulose.—Pine wood sawdust was treated by digestion for fourteen days with dilute nitric acid with addition of chlorate (Schulze). The mass was washed and digested with alkaline lye (1.25 p.ct. KOH), and exhaustively washed, treated with dilute acetic acid; again washed, and finally air-dried.

This product was found to yield 2.3 p.ct. furfural on distillation with HCl (1.06 sp.gr.).

(c) Purified wood.—Pine wood sawdust was treated in succession with dilute alkalis and acids, in the cold, and with alcohol and ether until exhausted of products soluble in these liquids and reagents.

In addition to the above the authors have also employed jute fibre and raw cotton wool in their investigations.

They note that the yield of cellulose is in many cases sensibly lowered by treating the material after drying at the temperature of 100°. The material for treatment is therefore weighed in the air-dry condition, and a similar sample weighed off