( 21 ) 
In a later pamphlet (Berlin, 1SS6,) von Forster again states 
that the gases generated by the detonation of the gun-cotton have, 
in the first instant, and as long as they exert their maximum force, 
the exact form and occupy the same space as was occupied by 
the gun-cotton before detonation and thus they produce an exact 
impression on the plate of the surface of the gun-cotton in con¬ 
tact with it, and he also says that the suddenness with which the 
power is exerted is shown by placing a leaf between the gun-cot¬ 
ton disk and the iron, for after detonation the whole frame or 
skeleton of the leaf will be found raised upon the iron. lie ex¬ 
plains that this is due to the larger as well as the smaller ribs of 
the leaf protecting the underlying parts of the iron while the 
thinner parts between could not yield such protection, and under 
them the impression is deeper. 
This was the condition of the subject when I again began to 
study it experimentally in 1SS6. I then first detonated gun-cotton 
disks in which the figures and letters were indented and obtained 
impressions on the plates, in which these were also indented. 
I next used disks having raised letters and figures and obtained 
impressions in which these were also raised. Next I cut deep 
channels in the disks, of various forms, taking care that they al¬ 
ways communicated with the outer air so that there would be no 
air confined in them, and again the impressions were exactly re¬ 
produced in the iron. I next filled the indented letters and fig¬ 
ures, in disks such as Jewell used, with paraffine and with vase¬ 
line, so that the material was flush with the surface of the disk, 
and, on detonation, the letters and figures were found to have 
been obliterated. I next struck, with stamps, similar letters and 
figures on an iron plate. This plate was laid face downwards on 
another iron plate and a lettered gun-cotton disk was placed on 
the upper plate and detonated. The result w’as, that while the 
gun-cotton disks produced the usual indented letters on the back 
of the top plate of iron, the top plate in whose letters and fig¬ 
ures air was also confined and which was subjected to the same 
blow, produced raised letters and figures on the iron plate on 
which it rested. These last three experiments certainly seem to 
prove that the resistance of air to compression has nothing to do 
with this action. Again, when we consider how enormous the 
