

1862.] 253 



of starch is added to blood. The corpuscles cohere more closely 

 and tenaciously than in either of the other conditions. They may 

 be said to blend with each other, inasmuch as they form homo- 

 geneous masses. If the solution of gum or gelatine be added to the 

 blood subsequently to the occurrence of the modes of arrangement 

 previously described, in such a manner as not to disturb them, the 

 already adherent corpuscles will cohere more closely, and the out- 

 lines of the corpuscles will be rendered less apparent, till at length 

 homogeneity of appearance results. The same effect takes place if a 

 drop of blood be allowed spontaneously to thicken, but rarely to such 

 an extent as on the addition of gum, gelatine, &c. 



This last mode of arrangement includes both the former condi- 

 tions, inasmuch as the corpuscles cohere firmly not only by their 

 biconcave surfaces but by their edges also. 



The first mode, or the formation of rouleaux, may be closely 

 imitated by preparing a number of small disks of cork so poised as 

 to float in the vertical position ; however carelessly these disks may 

 be thrown into water, they will be found in a few moments to have 

 arranged themselves into rouleaux after the manner of the blood- 

 disks. 



If a collection of blood-corpuscles cohering in the second mode 

 be compared with the manner in which bubbles group themselves, 

 as already described, the similarity will be allowed to be very 

 striking. 



From the marked analogy existing in their modes of arrangement, 

 a suspicion naturally arises that the blood-globules are influenced by 

 the same laws as the bubbles and cork-disks. A more critical investi- 

 gation, however, proves that the phenomena, although allied, possess 

 well-marked distinctions. The capillary action leading to adhesion, 

 as exhibited in the cork-disks and bubbles, is not possible if these 

 bodies are completely submerged in liquid ; but experiments carefully 

 performed with the blood-globules demonstrate that both the forma- 

 tion of rouleaux and the peripheral adhesion may take place under 

 circumstances in which it is absolutely certain that each individual 

 corpuscle is completely submerged. Thus if we take two pieces of 

 microscopic glass, and, placing them in contact, press them firmly 

 together and maintain them in opposition by a strong pair of pincers, 

 we shall still have a capillary space between them which will draw in 



