slowly in the sending of the message is the principal
stumbling-block which disconcerts ordinary telegraphic operators when
they come to try wireless telegraphy. For remedying this defect the
most hopeful outlook is in the direction of a multiplication of the
pieces of apparatus for spark-making and the combining of pairs of
them in such a way that, whenever the first one fails during an
appreciable interval of time to emit a spark, the second is called
into requisition. In this way a constant stream of sparks may be
ensured, without incurring the risk of running faster than the coil
will supply the electrical impulses necessary for the transmission of
the message.

Increased rapidity in land telegraphy by the ordinary system of
transmission by wire, and facility in making the records at the
receiving end in easily read typewriting--these are two desiderata
which at the close of the nineteenth century have been almost
attained, but which will take some time to introduce to general
notice. In the commercial system of the twentieth century the
merchant's clerk will write his messages on a typewriter which
perforates a strip of paper with holes corresponding to the various
letters, while it sets down in printing, on another strip, the letters
themselves. The latter will be kept as a record, but the former will
be taken to the telegraph office and put through the sending machine
without being read by the operator. The message will print itself at
the other end and wrap itself up in secret, nothing but the address
being made visible to the operator.

For the use of the general public who are not possessed of the special
apparatus necessary to perforate the paper another system is
available. Sets of movable type may be provided at the telegraph
office in small compartments, the letters being on one side and
indentations corresponding to the required perforations being cut or
stamped into the other sides of the movable pieces. The sender of a
message will set it up in a long shallow tray or "galley" like those
used by printers, and he will then turn the faces of the letters
downwards and see the whole passed through the machine without being
read by the operator; after which he can distribute the letters if he
chooses. In this way telegraphy will gradually become at once far more
secret and far cheaper than it is at present, and a large amount of
correspondence which at present passes through the post will be sent
along the