581 
angular forms to the crystalline form of hemoglobin, and 
suggest that they are possibly due to modified crystallisation. 
Then, thirdly, the corpuscle resumes an oval shape (fig. 
16 b, b), but not the original oval shape. It is now shorter 
and broader, and very distinct from the normal ellipsoid. Soon 
after this oval form has been assumed, numbers of the cor- 
puscles are observed to become colourless (fig. 16¢). There 
is no collapse, no movement on their part, but simply their 
colouring matter passes from them, and they remain as oval 
“‘chosts”’ or stromata of very definite shape with pellucid 
nucleus and well-marked outline. All this is under the in- 
fluence of chloroform vapour drawn into the gas chamber, as 
described. Acetic acid vapour now drawn in granulates the 
nucleus very sharply (fig. 16d). If instead of acetic acid 
vapour more chloroform vapour is drawn, no further result is 
obtained, the “ ghosts’ remain unchanged. But if, instead 
of the vapour, the observer now proceeds to make use of the 
liquid chloroform, vigorous action is obtained. A drop of 
chloroform placed on the already decolorised corpuscles 
causes an intense action, which looks like effervescence, 
around each ghostly stroma (fig. 170, 6). Hundreds of minute 
globules, the character of which is not gaseous, form along 
the sides of the stromata and along the edge of the nuclei, 
and then float off and disappear in the plasma. This action 
goes on with repeated additions of the chloroform until the 
outer portions of the stromata are dissolved, so that only the 
nuclei remain. These seem to be more resistant, but finally 
are broken up also with formation of the evanescent globules 
(eXtT ee; Ff): 
When chloroform liquid is added to fresh red blood- 
corpuscles of the frog, they almost instantly give up their 
hemoglobin, becoming nearly spherical, and the stromata 
have a collapsed and less regular outline than when chloro- 
form vapour is gradually allowed to act. 
Human corpuscles, subjected to the gradual action of 
chloroform vapour or to the action of chloroform liquid, ex- 
hibit identical appearances. 
15. Liffect of bisulphide of carbon.—With carbon bisulphide 
vapour I have obtained in the frog’s corpuscles nearly 
identical results with those given by chloroform—the assump- 
tion of an angular form (fig. 15 4), followed by a short oval or 
spherical form, the diffusion of the colouring matter, and the 
preduction of colourless stromata. It appears to take longer 
to discharge the colouring matter than chloroform, though the 
change of form is rapidly produced. The direct addition of 
the liquid to the corpuscles caused the solution of the stro- 
