On the Reduction of Observations. 135 



is placed upright in a wide-mouthed bottle of hard white flint- 

 glass about 10 centim. high and 5*5 centim. diameter, in which 

 about 50 to 70 grammes of paraffin-wax have been melted. 

 When the paraffin solidifies it contracts greatly ; but if the 

 bottle be not too large, holds the stem firmly in its place. To 

 keep out the dust a lid, formed from a disk of gutta-percha 

 about 2 millim. thick, softened by dipping into boiling water, 

 is placed loose-tight on the stem. The arrangement is shown 

 in fig. 1. The upper end of the central tube is open, and 

 affords a convenient means of placing on the support various 

 different objects, such as a metal ball fixed on a metal rod, or 

 a flat metal plate on which to stand any object that is to be 

 insulated. For carrying wires over a table in experiments 

 requiring high insulation, rods of flint-glass of 25 to 30 centim.' 

 length, curled at the top in the form of a crook or other hook- 

 form, as shown in figs. 2 to 5, are placed in the central tube 

 of the insulating support, and the wires are slung in them. 

 The insulation-resistance of these supports is many hundreds 

 of megohms even in damp weather. Should a film of dust 

 accumulate on the surface of the paraffin in consequence of any 

 neglect in lowering the gutta-percha cap, the insulator only 

 requires to be warmed to the melting-point of paraffin to 

 restore the lost insulation. For work requiring very special 

 insulation, sulphuric acid is poured over the top of the paraffin; 

 and in some cases solid stems of glass have been used instead 

 of glass tube. Hitherto there has been no trouble from yield- 

 ing in the paraffin, which was feared at the outset as a possible 

 fault. A good hard paraffin has been used ; and as the flat- 

 tened form given to the basal enlargement of the glass tube 

 constitutes in itself a foot to the central support, there is little 

 or no tendency for the paraffin to bend under the weight that 

 may be placed for a few hours on the top of the support. To 

 add stability to the apparatus, the bottle is let into a wooden 

 foot. The total cost of the apparatus is less than one tenth of 

 the price charged for the apparatus of Mascart. 



* 



XVII. On the Reduction of Observations. By F. Y. Edge 

 WORTH, M.A.yLecturer on Logic at King\s College, London" 



IN an article on the Method of Least Squares which lately 

 appeared in this Journalf I attempted to answer two 

 questions — first, ichy is one mode of reducing observations to 

 be preferred to another, and, secondly, what is the preferable 

 mode in certain important cases. The answer to the first 



* Communicated by the Author. 

 t November 1883. 



