72 CRUCIFER^. 



often almost cylindrical, throwing off here and there filiform and long 

 slender cylindrical roots, and finally dividing into two or three branches. 

 The root is of a light yellowish brown ; internally it is fleshy and 

 perfectly white, and has a short non-fibrous fractui-e. Before it is 

 broken it is inodorous, but when comminuted it immediately exhales 

 its characteristic pungent smell. Its well-known pungent taste is not 

 lost in the root carefully di'ied and not kept too long. 



A transverse section of the fresh root displays a large central 

 column with a radiate and concentric arrangement of its tissues, which 

 are separated by a small greyish circle from the bark, whose breadth is 

 from -5- to 2 lines. In the root branches there is neither a well-defined 

 liber nor a true pith. The short leaf-bearing branches include a large 

 pith surrounded by a circle of woody bundles. The bark adheres 

 strongly to the central portion, in which zones of annual gi-owth are 

 easily perceptible, at least in older specimens. 



Microscopic Structure — The corky layer is made up of small 

 tabular cells as usual in suberous coats. In the succeeding zone of 

 the middle bark, thick-walled yellow cells are scattered through the 

 parenchyme, chiefly at the boundary line of the corky layer. In the 

 root the cellular envelope is not strikingly separated from the liber, 

 whilst in its leafy branches this separation is well marked by wedge- 

 shaped liber bundles, which are accompanied by a group of the yellow^ 

 longitudinally-elongated stone cells. The woody bundles contain a few 

 short yellow vessels, accompanied b}'- bundles of prosenchymatous, not 

 properly woody cells. The centre, in the root, shows these woody 

 bundles to be separated by the medullary parenchyma ; in the branches 

 the central column consists of an uniform pith without woody bundles, 

 the latter forming a circle close to the cambium. The parenchyma 

 of the whole root collected in spring is loaded with small starch 

 gi-anules. 



Chemical Composition — Among the constituents of horse-radish 

 root (the chemical history of which is however far from perfect) the 

 volatile oil is the most interesting. The fresh root submitted to dis- 

 tillation with water in a glass retort, yields about -J- per mi He of oil 

 which is identical with that of Black Mustard as proved in 1843 by 

 Hubatka. He combined it with ammonia and obtained crystals of 

 thiosinaramine, the composition of which agreed with the thiosinammino 

 from mustard oil. 



An alcoholic extract of the root is devoid of the odour of the oil, 

 but this is quickly evolved on addition of an emulsion of White Mustard. 

 The essential oil does not therefore pre-exist, but only sinigrin 

 (myronate of potassium) and an albuminoid matter (myrosin) b}'^ whose 

 mutual reaction in the presence of water it is formed (p. 6Q). This 

 process does not go on in the growing root, perhaps because the two 

 principles in question are not contained in the same cells, or else exist 

 together in some condition that does not allow of their acting on each 

 other, — a state of things analogous to that occurring in the leaves of 

 Laurocerasus. 



By exhausting the root with water either cold or hot, the sinigrin 

 is decomposed and a considerable proportion of bisulphate is found in 

 the concentrated decoction. Alcohol removes from the root some fatty 



