lamp a peculiar, purplish light. "This is a machine for the
propagation of ultra-violet rays."

He placed the second slide, with its germs of anthrax, in the light,
allowing it to play over the slide.

"Now look," he said.

We did. Something had evidently happened. The chains were broken and
smaller units were moving.

"If anthrax germs are exposed for a few seconds, even, to ultra-violet
light, they change more or less," Kennedy proceeded. "These new forms
are not stable. They quickly change back again into their original
form."

For about ten minutes we sat in silence while the weird light played as
if with ghostly fingers on the deadly invisible peril on the little
glass microscope slide.

"But if the action of the ultra-violet rays is continued," went on
Craig, "the microbe changes into a coccus, and then into a filiform
bacillus. This form is stable. And the germ is changed in other
respects than mere shape. It has entirely new characteristics. It is a
true mutation. It produces a disease entirely different from that of the
anthrax bacillus from which it is derived. I have tried it on a guinea
pig--and it has died in forty-eight hours."

Startled as I was by this remarkable discovery, I yet had time to watch
Haynes. He had not taken his eyes off Kennedy once since he began to
speak.

"In anthrax," continued Craig, "an autopsy reveals an enormous serous
swelling, about the point of inoculation, with a large gathering of
microbes which are formed in the blood and spleen. Death seems to be due
to septic poisoning. But this new microbe--super-toxicus, I think it
might well be named--produces no swelling and scarcely any microbes are
to be found in the blood.

"The lungs are inflamed, with innumerable small abscesses. That is the
internal form of the disease from breathing in the spores of these
microbes. It has an external form, also, but that is by no means so
deadly. One would say that death from the internal form of the disease
was due to poisoning. The toxin of this microbe produces a sort of
asphyxiation, cutting off and eating up the supply of oxygen.

"Such a condition is called cyanosis. That is why in Delaney it had the
appearance of cyanogen poisoning. The effect was the same. But the trace
of cyanogen in the air was merely a coincidence, Haynes. It wasn't
cyanogen that killed. But it was something quite as deadly--and far
harder to trace--a new germ!"

We listened, fascinated.

"A French sc