1870.J Physics. 135 



having been exposed to sunlight, the best and whitest glass made 

 at St. Gobain is turned very distinctly yellow ; extra white glass 

 (of a peculiar mode of manufacture) becomes even more yellow, and 

 gradually assumes a colour known as pelure d'oignon ; glass con- 

 taining 5 per cent, of litharge is also affected, but far less percep- 

 tibly ; crystal glass, made with carbonate of potassa (the other 

 varieties referred to contain carbonate of soda), litharge, and silica, 

 is not at all affected ; English plate-glass, made by the British 

 Plate-Glass Company, and exhibiting a distinctly azure-blue tinge, 

 remains also unaffected. The author attributes this colouration, 

 which begins with yellow and gradually turns to violet, to the 

 oxidizing effects of the sun's rays upon the protoxides of iron and 

 manganese contained in glass. 



A solution of iodide of potassium is, even when kept in well- 

 closed bottles, slowly decomposed by the action of daylight, and 

 assumes a somewhat yellowish tinge due to free iodine. M. Loew 

 filled a number of glass tubes for about from one-half to three- 

 fourths of their capacity, with a solution of iodine of potassium, 

 and, after having sealed these tubes, exposed them to direct sun- 

 light. Another set of tubes were likewise filled with the same 

 solution, but all air was expelled, and the tubes sealed during and 

 after the solution had been boiling for a considerable time. These 

 tubes were also exposed to the action of direct sunlight ; after 

 three or four months' exposure the tubes and contents were 

 examined, those wherein no air at all was left were found to be 

 perfectly colourless, no decomposition of the contents having taken 

 place. As regards the other tubes, the following results are 

 noticed : — (1) Under the influence of light, the oxygen of the air 

 decomposes iodide of potassium, iodine in small quantity is set free, 

 while hydrate of potassa is found in the liquid. (2) This decom- 

 position is limited, and does not even, when a large quantity of 

 oxygen is present, increase, because a portion of the iodine set free 

 enters again into combination with the caustic potassa set free, 

 forming iodide of potassium and iodate of potassa. (3) The testing 

 for ozone by means of a solution of iodide of potassium and starch 

 (or paper prepared therewith) is of no value whatever, unless care 

 has been taken to exclude direct sunlight. 



Heat. — M. Dufour recommends the following process for 

 demonstrating that the flame of a candle is formed of a hollow 

 cone, luminous on the outside only, and dark in the interior. A 

 caoutchouc tube has, at one of its extremities, a gas jet with an 

 almost semicircular slit of . 04 metre in depth. The other end of 

 the tube communicates with a reservoir of water placed at a con- 

 venient height. Upon a suitable pressure, the water flows out by 



