274 
This ellipticity of the blood-corpuscles of the lower vertebrates 
is, moreover, again adaptive, since it is well known that in travers- 
ing the smallest capillaries in such forms as the frog, for example, 
every corpuscle must in traversing the latter place its long axis 
parallel, or nearly so, with the axis of the lumen of the vessel in 
order to pass through; in other words, the least diameter of the 
corpuscle only can traverse the lumen of capillaries. In thus ad- 
justing its long axis to the axis of the lumen of the capillary, sucha 
corpuscle again presents a maximum amount of surface and volume 
in the nearest relation to the tissues it is to oxidize, whereas if it 
traversed the capillary with its long diameter transverse to the axis 
of the lumen of the capillary, it would present the least amount of 
its volume and surface in the nearest relation to the tissue to be 
oxidized. Were this last condition prevalent it would require 
many more corpuscles to do the same work as is done by the ellipti- 
cal corpuscles that now traverse the capillaries lengthwise, or with 
_their long axes parallel with the axis of the lumina of the latter. 
A further supposition that may be made regarding the elliptical 
blood-corpuscles is that, originally globular, they became at first dis- 
coidal, then elliptical, and that the elliptical form of the disks was 
consequent upon slight constraint within the capillaries during their 
passage through the latter ; in other words, the elliptical form was de- 
rived from the discoidal as the discoidal was derived from the still 
more primitive globular form. We may also suppose that very 
slight mechanical constraint in passing through the capillaries 
would distinctly tend to develop a tendency towards converting the 
disks that were slightly too large to pass through the capillaries into 
elliptical disks. JI can thus conceive a mechanical origin for the 
elliptical form of the red blood-corpuscle in general wherever it 
occurs. 
It may also be assumed that the size of the corpuscle is directly 
related to the rate of metabolism of the organism. Thus in the slug- 
gish batrachia the corpuscles are large ; in the more active fishes and 
reptiles smaller ; in mammals still smaller, and in the most active 
mammals and birds, such as the musk deer and humming bird, 
smallest of all, reaching minimum dimensions of zo/55 tO goog Of 
an inch in these types, according to Gulliver. 
The impulse that tends to develop a vortical flux of matter from 
opposite poles of a young globular red blood-corpuscle in every 
direction from the centre, it would be difficult to specify further 
