SCIENTiriC SUMMARY. 
331 
is small, burning only about 1|- or 2 cubic feet per hour j but be has a larger 
one in course of manufacture, which he expects will be infinitely more sen- 
sitive. 
Peculiarities of Spectrum Analysis.— In. passing the induction spark 
between two metallic pieces, M. L. de Boisbaudran says that the spectral 
lines are chiefly or exclusively obtained at the negative pole ; but with 
aluminium the contrary occurs. Certain solutions examined with the spark 
only give the air spectrum, even when negative ; by moistening the exterior 
platinum wire this may be avoided. The presence of metallic molecules in 
the interpolar space enfeebles the air spectrum. The exterior wire may be 
brought nearer the solution when the latter is positive than when it is 
negative, in the latter case the liquid tends to meet the wire. The projec- 
tion of the liquid in droplets ; its pulverisation increases so rapidly with the 
length of spark, and varies with the liquid. The dilution of liquors in- 
fluences the relative intensities of the lines. When, after long calcination 
in a gas-flame, there is hardly a trace of matter on the platinum wire, it is 
not always the lines originally brightest which persist longest. With the 
spark many metals do not give lines when in the compact state, but furnish 
beautiful spectra when finely divided. It would sometimes be of advantage 
to mix the metals with pure lead, from which the spark should be brought. 
— Comptes Rendus,” May 12th. 
A Neiu Electric Insulator. — Mr. H. Highton writes to the editor of the 
Chemical News” (May 2nd) that some time ago he discovered, in the 
course of experiments, that vegetable tar, by the addition of the oxides of 
lead (or in a less degree by some other substances), is almost instantly 
solidified, and that the solid substance thus obtained has remarkable insu- 
lating powers. In some experiments at Silvertown he found that No. 18 
copper wire, covered with a coating of gutta-percha weighing only 21 lbs. 
to the mile, had its insulation increased nearly 200,000 per cent., and that 
the insulation resistance was no less than 2,800,000,000 units, an insulation 
sufficient for any lengths possible on the surface of the earth. Insulation of 
so great cheapness makes all further experiments on telegraphy without in- 
sulation needless, though in fresh water or by land this can be effected to 
very great distances. 
Preventing Ship's Compasses being affected hy Iron Cargoes. — A new method 
for this purpose has been devised by M. Gloessner, and is described in a 
paper which he has published in the “ Bulletin de TAcademie Royale des 
Sciences, etc. etc. de Belgique,” No. 4, 1873. 
Experiment in Electro-Dynamics with a Gramme Magneto-Electric Ma- 
chine . — The following experiment with this machine is described in Comptes 
Rendus ” by MM. Plante and Niaudot-Breguet. It is of special interest 
to the reader, because this machine is now described in our pages : — If a 
secondary couple (of lead plates) be charged with a Gramme magneto- 
electric machine, and then be left undischarged in connexion with the 
machine which has been stopped, the latter presently begins to turn under 
influence of the current from the secondary couple in the same direction as 
it was turned in charging ; and continues turning (more slowly, indeed, yet 
at a considerable rate) for two or three minutes till the couple is discharged. 
In this restitution there is, probably, only slight loss. That the machine in 
