ould be self-registering. Each was obliged to design his
own, being dissatisfied with the photographic registers commercially
available. The development of these systems would therefore appear to
have been due, in part, to the general spread of a conviction that
satisfactory instruments were attainable.

[Illustration:

A, is the Vane.

B, is the Perpendicular Shaft.

C, is a Horizontal Circular Plate of light material attached to the
shaft.

E and F, two Rollers communicating motion to the Apron E F from left to
right.

1, 2, 3, &c., are minute Cards, placed upon the Apron.

G, is a Clock that regulates the motion of the Roller E, and
consequently that of the apron and cards.

D, is a small weight to relieve the Clock.

N, NE, E, &c., are paper boxes placed upon the circular plate, to
receive the cards, as they fall from the apron at E.

Figure 7.--In 1838 the pioneer American meteorologist James H. Coffin
(1806-1873) devised a self-registering wind direction indicator; in 1849
he improved it as shown here. The band, moved by clockwork, carries
cards marked with the day and hour. In Coffin's earlier instrument, a
part of which is now in the Smithsonian Institution, the vane carried a
funnel for sand, which ran into a circular row of bottles. (From
_Proceedings of the American Association for the Advancement of
Science_, 1849, vol. 2, p. 388.)]

This confidence was warranted, for the decade of the 1850's had seen the
appearance of major innovations in the basic instruments--thermometer,
barometer, and wind velocity indicator--that made available instruments
more adaptable to self-registration. It also saw the development of a
new method of electrical registration derived from the telegraph. Sir
Charles Wheatstone initiated this small revolution in 1843 when he
reported to the British Association that he had constructed an
electromagnetic meteorological register which "records the indications
of the barometer, thermometer and the psychrometer [meaning wet-bulb
thermometer] every half hour ... and prints the results on a sheet of
paper in figures," running a week unattended. The working of this
register involved the insertion of a conductor in the tubes to make a
circuit, the thermometers having open tops.[26] This was ten years after
the development of the electromagnetic relay and six years after
Wheatstone's introduction of his own telegraph.

Wheatstone's instrument left a very ephemeral record in the