772 : REPORT—1890. 
2. Report of the Committee on the Silent Discharge of Electricity in Gases. 
See Reports, p. 338. , 
3. Report of the Committee on the present Methods of Teaching Chemistry. 
See Reports, p. 265. 
4. On Recent Legislation as Facilitating the Teaching of Science. 
By Sir Henry Roscon, MP., F.B.S. 
5. The Refraction and Dispersion of Fluorbenzene and Allied Compounds. 
By J. H. Guavstonz, Ph.D., F.R.S., and Guorce GuaDstTone. 
The authors had determined the molecular refraction of very pure specimens of 
fluor-, chloro-, bromo-, and iodo- benzene for the solar lines ACDEGH. The 
compounds of the three more common halogens gave the following results for 
chlorine, bromine, and iodine :— 
Chlorine Ry = 10:00, Ry_, = 0-70 
Bromine ,, =15:28, ,, =1°41 
Jodine. ,, =25:20, ,, =3'43 
These are in conformity with numbers previously determined for the halogens, 
especially when deduced from such bodies as bromoform, dibromide of ethylene, 
&c. In the case of the fluoride the molecular refraction is exceedingly small, and 
smaller for each successive line of the spectrum; so that for 
Fluorine Ry = 4+ 0°63, Ry-a = —028 
This small molecular refraction is also in accordance with what was previously 
known ; but a negative dispersion of the same character has never before been 
observed. Upon examining, however, the refraction of fluorspar and aqueous 
solutions of fluoride of potassium, the fluorine in them was found to exhibit the 
same reversal. The double fluosilicates examined by Topsoe and Christiansen also 
appear to lead to the same inference, though the data for exact calculation are 
wanting. 
6. A Method of Quantitative Analysis. By G. H. Bary, D.Sc., Ph.D., 
and J. C. Carn. 
The method consists in precipitating in the ordinary manner, and weighing the 
precipitate in the liquid, having previously determined its specific gravity. The 
specific gravity of the liquid and of the precipitate being known, and the volume of 
the flask in which the weighing is made, it is possible to calculate the weight 
of the precipitate directly, avoiding the troublesome operations of filtering and 
washing. 
It is not necessary to wash the precipitate free from the supernatant liquor, since, 
by having two flasks of, say, 100 c.c. content, and filling one with the supernatant 
liquor, and the other with supernatant liquor and precipitate, and then determin- 
ing the weight of both, we have all the data required. The 100 c.c. flasks as 
ordinarily made for volumetric analysis are too wide in the neck to admit of 
accuracy, and flasks having narrow necks with graduations were used in the 
experiments. Special arrangements were also made to overcome the difficulty of 
introducing the liquids and precipitate through a neck of such small diameter. 
The method is specially recommended for commercial analyses where a 
tolerably large quantity of the sample to be determined is available, and the 
amount taken should be such as to yield not less than 5 grammes of precipitate. 
A manifest objection that might be raised to such a process is that the specific 
gravity of the precipitate varies according to the circumstances under which precipi- 
tation occurs. ; 
