TBANSACTIONS OP SECTION B. 109 



the tube. There was a strong escape of gas on opening the 

 tube. On heating, the white mass gave a colourless volatile 

 liquid with a penetrating onion odour, which formed white 

 flakes of magnesium oxide on entry of the smallest trace of air, 

 and evolved thick white fumes when warmed in air, but did not 

 spontaneously inflame. 



This liquid consisted apparently of free hydrocarbons with 

 traces of magnesium ethide. The greater part of the latter was 

 believed to be combined with magnesium iodide in the white 

 residue, for this possessed the property, even after prolonged 

 and strong heating, of decomposing water with explosive force, 

 evolving heat and a penetrating smelling gas. 



Cahours proceeded in a similar way. On adding the ethyl 

 iodide to the magnesium filings, so much heat was evolved that 

 the tube had to be cooled with water. The latter was then 

 sealed, and heated up to 120°— 130°, till all the liquid dis- 

 appeared. The white residue was then transferred to a retort 

 filled with hydrogen, and distilled in a current of that gas, when 

 a liquid was obtained which took fire in the air, and acted 

 strongly on water. This was partially fractionated ; and on 

 analysis two fractions gave 54"5 and 55*2 per cent, carbon, and 

 11*4 and 11-5 per cent, hydrogen (calculated for MgEto, — 

 carbon 58-53, hydrogen 12-19). The error was attributed to 

 the small amount of material available, which did not admit of 

 sufficient fractionation ; and it may be presumed that the same 

 cause accounts for the absence of analytical proof of the pre- 

 sence of magnesium. 



"Wanklyn treated sodium zinc ethyl with mercury and 

 magnesium wire. A white solid was formed, which was spon- 

 taneously inflammable in air, and which contained magnesium 

 and zinc, but only traces of sodium ; while on the other hand 

 the mercury was found to be rich in sodium. The white solid 

 was believed to be magnesium zinc ethyl, formed according to 

 the following equation : — 



Hg+Mg+SNaC^Hs . Zn(C2H5).2 = HgNa,-l-Mg(CoH5),Zii(CjH5)o. 



But the experiment was merely a qualitative one. 



The magnesium used by me was the ordinary ribbon of 

 commerce, in which I failed to detect any impurity. In order 

 to get it in a finely divided state, I first tried distilling it in 

 hydrogen. For this purpose an apparatus was constructed, 

 consisting of a lin. iron tube about 18iu. in length, fitted b}^ 

 means of a reducing-socket to a 2in. tube, 6iu. in length, to act 

 as a receiver ; and to this again another short piece of lin. 

 tube was attached. To each end of the apparatus about a foot 

 of i-in. iron tube was fitted, for the convenience of connecting 

 the hydrogen generator. The magnesium was placed in the 

 longer lin. tube, and the apparatus was screwed together, the 



