332 BULLETIN OF THE 



of the mass, and are therefore the first to commence breaking- 

 apart. Accordingly on ultimate separation, each fragment ex- 

 hibits a hollow or cup-like surface of fracture, where the interior 

 portion of the material has first parted : the depth of the con- 

 cavity being somewhat proportioned to the malleability of the 

 substance. " With substances of greater rigidity, this effect is 

 less apparent, but it exists even in iron, and the interior fibres 

 of a rod of this metal may be entirely separated, while the outer 

 surface presents no appearance of change. From this it would 

 appear that metals should never be elongated by mere stretching, 

 but in all cases by a process of wire-drawing, or rolling. A wire 

 or bar must always be weakened by a force which permanently 

 increases its length without at the same time compressing it."* 



Hydrometric Experiment. — A novel project for the rectifica- 

 tion of spirits by the simple process of static separation of the 

 alcohol and water by the stress of their specific gravities when 

 exposed in long columns, produced in 1854 a considerable sensa- 

 tion. It was alleged that the coercitive compression exerted by 

 the water in a long hydrostatic column greatly accelerated the 

 displacement and separation induced by gravitation, and that 

 only a few hours were necessary to complete the process, if the 

 column were sufficiently high.t 



A patent was obtained : affidavits and samples fully attested 

 the wonderful efficiency of the process ; and only the co-operation 

 of confiding capitalists was required, to realize fabulous profits, 

 and effect a manufacturing and commercial revolution. 



Simply in the interests of truth, Henry undertook the careful 

 investigation of this surprising pretension. One of the towers of 

 the Smithsonian Building supplied a convenient well for the 

 experiment, easily accessible throughout its height. "A series 

 of stout iron tubes of about an inch and a half internal diameter 

 formed the column ; the total length of which was one hundred 

 and six feet. Four stop-cocks were provided ; one at the bot- 

 tom, one about four feet from the top, and the other two to the 

 intermediate space equally divided or nearly so." Very careful 

 hydrometer and thermometer registers were made at increasing 

 intervals of time, the last being that of nearly half a year: a 

 portion of the reserved liquor being simultaneously tested. The 

 result stated, is : " There is not the slightest indication of any 



* This conclusion is not at all in opposition to the ascertained fact of 

 the increased strength imparted to an iron rod by thermo-tension, dis- 

 covered by Prof. Walter R. Johnson. 



f An incidental remark in Gmelin's " Handbook of Chemistry" seemed 

 to give some color of plausibility to the scheme. " Brandy kept in casks 

 is said to contain a greater proportion of spirit in the upper, and of water 

 in the lower part." Gmelin's Handbook, Translated by Henry Watts. 

 London, 1841, part i. sect. 4, — vol. i. p. 112. 



106 



