the machines of Holmes, of
the Alliance Company, of Siemens, and of Gramme, were pitted against
each other. The Siemens and the Gramme machines delivered direct
currents, while those of Holmes and the Alliance Company delivered
alternating currents. The light of the latter was of the same
intensity in all azimuths; that of the former was different in
different azimuths, the discharge being so regulated as to yield a
gush of light of special intensity in one direction. The following
table gives in standard candles the performance of the respective
machines:

Name of Machines. Maximum. Minimum.

Holmes 1,523 1,523

Alliance 1,953 1,953

Gramme (No. 1). 6,663 4,016

Gramme (No. 2). 6,663 4,016

Siemens (Large) 14,818 8,932

Siemens (Small, No. 1) 5,539 3,339

Siemens (Small, No. 2) 6,864 4,138

Two Holmes's coupled 2,811 2,811

Two Gramme's (Nos. 1 and 2) 11,396 6,869

Two Siemens' (Nos. 1 and 2) 14,134 8,520

[Footnote: Observations from the sea on the night of November 21,
1876, made the Gramme and small Siemens practically equal to the
Alliance. But the photometric observations, in which the external
resistance was abolished, and previous to which the light-keepers had
become more skilled in the management of the direct current, showed
the differences recorded in the table. A close inspection of these
powerful lights at the South Foreland caused my face to peel, as if it
had been irritated by an Alpine sun.]

These determinations were made with extreme care and accuracy by Mr.
Douglass, the engineer-in-chief, and Mr. Ayres, the assistant engineer
of the Trinity House. It is practically impossible to compare
photo-metrically and directly the flame of the candle with these
sun-like lights. A light of intermediate intensity--that of the
six-wick Trinity oil lamp--was therefore in the first instance
compared with the electric light. The candle power of the oil lamp
being afterwards determined, the intensity of the electric light
became known. The numbers given in the table prove the superiority of
the Alliance machine over that of Holmes. They prove the great
superiority both of the Gramme machine and of the small Siemens
machine over the Alliance. The large Siemens machine is shown to
yield a light far exceeding all