68 Transactions of the Royal Microscopical Society. 
taken directly from the living animal. They represent minute 
colourless rods of yoW to yo%o thickness, and from to 
y^f 0 mm. in length ; they are slightly indentated, giving them 
the appearance of being composed of a number of segments (Fig. 15). 
In fact, they so closely resemble Bacteria, that when I first saw 
them I was impressed with the idea that they were dead specimens 
of these ambiguous beings. But I soon became convinced of their 
crystalline nature, by discovering that they dissolved when a drop 
of water was added to the preparation. 
Not unfrequently blood-corpuscles are met with in the fresh 
blood, containing a considerable number of larger or smaller 
vacuoles in their interior. They resemble ordinary air-bubbles, 
being distinguished by a dark contour, which is gradually lost in a 
slight shade and a perfectly clear centre, and exhibits a light 
greyish tint, playing into violet (Fig. 16). In some cases the 
peripheral portion of these vacuoles is light, and of a pinkish tint, 
though with a dark contour, while their central portion is slightly 
shaded (Fig. 17). I have no idea of the exact nature of these 
bodies. As I found them especially numerous in the blood-cor- 
puscles of an Amphiuma, which I had left three months without 
food, I am inclined to think that their formation is owing to some 
decomposition of the protoplasm. They invariably, disappear on 
the application of water. 
Having thus far described the coloured blood-corpuscles of the 
Amphiuma as they are met with in the fresh blood, together with 
the different phenomena which they exhibit in this condition, we 
will turn our attention to their behaviour when acted on by 
different reagents, and begin with the most neutral of these, namely, 
water. 
When water is added to a specimen of fresh blood, the coloured 
blood-corpuscles will be observed to become gradually paler, until 
finally they are rendered perfectly colourless. This change, of 
course, is caused by the endosmotic current of the water penetra- 
ting into the blood-corpuscle and dissolving its colouring matter, 
which in consequence escapes into the surrounding liquid. 
During this process the blood-corpuscles do not swell ; on the 
contrary, many of them become actually smaller by the continued 
action of water. A number of tliem assume a round form, which 
however, is not permanent, for the greater part of them eventually 
resume their original oval form, or very nearly so. 
While the main body of the blood-corpuscle is rendered colour- 
less by the action of the water, certain changes are observed to take 
place in its nucleus. The granules contained in the interior of this 
body are dissolved, and, after first expanding, run into each other, 
representing a homogeneous mass of a pale greenish tint. At the 
same time the whole body of the nucleus, now distinguished by a 
