Reluctantly I accepted from the Colonel an invitation to dinner, for
the feeling was still strong in me that some danger was impending.
Half-way through dinner there came the well-known scream of an
approaching shell, which burst at the other end of the orchard. A
second shell burst a little closer; a third came closer still, and a
fourth rained shrapnel on the roof; all the others, with one
exception, fell short, and the shelling was over for the time. It was
just another one of those "intuitions."

While the shells were flying we all kept on eating as if this were a
usual everyday accompaniment to lunch, though I noticed that they
watched me with as much interest as I eyed them during the process,
each curious to know how the other took it.

The varied nature of laboratory work in the army and its practical
applications may be seen from the following examples:--

One day the O.C. of a hospital sent over a pint of tea suspected of
poisoning 28 out of 29 men who drank it. From the history of the
affair we did not believe that this could possibly be the cause, and
after making a few rapid tests to exclude metals, we proved that the
tea was not poisonous by the simple, practical test of drinking it,
Major Rankin being the official tester. This method of making a
practical physiological test rather astonished the British
authorities.

A German gas mask found on the battle field was submitted to us to
find what chemicals were present. That mistakes were sometimes made by
the Germans was evident when we found that the mask had not been
treated with chemicals at all; some of the Huns at least had been
unprepared for a gas attack.

The clarifying apparatus on the British water carts was mechanically
defective and usually broke at certain definite places.
Recommendations were made by us after we had experimented with rubber
instead of rigid connections, which resulted in all the water carts in
the British army being equipped with rubber connections, the results
being entirely satisfactory.

A great deal of experimental laboratory and field work was done with
chlorine gas and the efficiency of gas masks and helmets. Experimental
physiological and pathological work was done on animals with chlorine
and other gases, and on the drying out and deterioration of gas
helmets and the chemicals used in them. Subsequently a Gas Service was
inaugurated and all work of this sort carried out in special
laboratories at G.H.Q.

Quite