377 
the corpuscles almost instantly assume the form of small 
spheres. ‘Then it may be seen that the nucleus breaks up 
into bits which float about in the sphere (fig. 14 a), swimming 
round and round as the action goes on, then suddenly the 
sphere collapses, the colouring matter is diffused all round, 
and no trace of the corpuscle, except a few scattered specks, 
remain. ‘The complete disappearance of the corpuscle, or 
the survival of a ghost-like stroma, depends on whether a 
strong or a moderately intense solution of the ammonia be 
used. As I set to work to repeat the observations on the 
effect of weak ammoniacal gas, in the early spring, I was 
surprised at not at once getting the change in the shape of 
the corpuscle which I have seen in the summer. The 
results which follow are explicable on the supposition that 
the outer portion or wall of the corpuscle is denser and more 
resisting in the early spring than later. It has often been 
remarked that one cannot at will reproduce changes in form 
caused by reagents acting on blood-corpuscles, and this, it is 
most probable, is due to some slight variation in their con- 
stitution from season to season, and from individual to indi- 
vidual. The elongated, pointed, and triangular shapes 
drawn in fig. 9 a. I did again get this summer with very weak 
vapour. In the spring, however (as also in some cases this 
summer), the action of very weak ammoniacal gas on the 
frog’s blood gave three different types of action, which did 
not occur simultaneously im the same drop, nor lead one 
into the other, but seemed to depend on very slight differ- 
ences in the rate at which the ammonia was allowed to act, 
its strength, and the condition of the corpuscles themselves. 
The first condition which was obtained most frequently, 
and which I heve again obtained during the summer, is 
drawn in fig. 9 6. As soon as a very small amount of am- 
monia is drawn into the gas chamber, a very weak solution 
of the gas being used as the source, and air being allowed to 
bubble through it and thence into the gas chamber, the cor- 
puscles become irregular in shape, and assume lobular forms. 
The lobules tend to constrict themselves in various ways, and 
send out long irregular processes as represented in the figure 
(fig. 12), whilst the nucleus-remaining pellucid enlarges some- 
what. Acetic acid vapour of maximum strength now sub- 
stituted for the ammonia produces a remarkable result. The 
processes are withdrawn, and the corpuscles remain of an 
irregular form, but instead of the nucleus becoming granu- 
lated it remains perfectly clear and pellucid and much swollen, 
whilst the body of the corpuscle is granulated (fig. 12 a’). The 
matter which 1s precipitable by acetic acid passes under the 
