W. W. WAGNER, L. P. LATHAM, AND D. R. HALSEY 
255 
an L-W motion analysis projector. The individ- 
ual sequences are spliced so as to form an end- 
less loop and projected at a magnification of 
500X onto grid paper calibrated such that each 
grid line is equivalent to 50fx apart. The ob- 
server traces the perfused capillaries (defined 
as capillaries in which blood flows, as opposed 
to oscillates, at any time during the sequence). 
Every time a traced capillary touches or 
crosses a grid line, it is counted. The total 
counts are taken as an indication of the per- 
fused capillaries for that field. Because the 
same field is photographed during control and 
hypoxia, the grid counts provide a valid com- 
parison of the active capillaries under each con- 
dition. 
RESULTS 
Successful studies were made in 8 dogs : in 4 
animals the top of the lung v^^as studied and in 4 
dogs we studied the bottom of the lung. These 
represent about half of the total number of ex- 
periments. Most of the failures were due to the 
visceral pleura being too thick for satisfactory 
observations (generally found in older animals) 
or because the observation field moved too 
much. 
The animals tolerate the procedure well. 
Their blood-gases are normal during the control 
period (see Table I). The low arterial pH is 
probably caused by the anesthetic. However, for 
the purposes of this work, the acidemia serves 
to accentuate the hypoxia vasopressor 
response. ^"'ii There is no evidence of shock as 
indicated by blood pressures. 
There is a brisk rise in mean pulmonary ar- 
tery pressure (Ppa) with hypoxia: from a con- 
trol of 16 torr ± 0.7 (SEM) to 23 ± 1.1 torr 
Table I. — Response of Animals to Hypoxia 
Control 
Hypoxia 
Difference 
F pa (torr) 
16 ± 0.7 
23 ± 1.1 
p < 0.01 
Cardiac output 
(1/min) 
1.72 ± 0.09 
1.82 ± 0.09 
n.s. 
Sa02 (%) 
97 ± 0.3 
73 ± 1.4 
p < 0.01 
Pa02 (torr) 
117 ± 3.7 
51 ± 1.9 
p < 0.01 
PaCOs (torr) 
39.7 ± 1.8 
38.9 ± 0.9 
n.s. 
PH. 
7.27 ± 0.07 
7.28 ± 0.01 
n.s. 
resp. rate (per min) 
14.4 ± 1.9 
41.6 ± 3.4 
p < 0.01 
Temp. — esophageal (C) 
38.3 ± 0.3 
38.1 ± 0.3 
n.s. 
Values are means ± SEM 
n.s. (not significant), i.e., p > 0.05 determined by t-test 
with hypoxia (p < 0.01). Cardiac output rose 
in 12 cases with hypoxia and fell in 5 cases (see 
Figure 3). Arterial PCOo is essentially con- 
stant during control and hypoxia (see Table I). 
Examples of capillary tracings are shown in 
Figure 1. Under all conditions there is greater 
perfusion at the bottom of the lung compared 
with the top. Hypoxia causes obvious recruit- 
ment of new capillaries at the top of the lung. 
Hypoxia also causes recruitment at the bottom 
of the lung but the new capillaries are less ob- 
vious by visual inspection because the bed is al- 
most completely perfused during normoxia. 
The capillary grid count data are shown in 
Figure 2 and summarized in Table II. These are 
Table II.- 
— Capillary Counts 
(Recruitment) 
During 
Hypoxia 
Control 
Hypoxia 
Difference 
Top 
53 ± 7 
92 ± 9 
p < 0.01 
Bottom 
153 ± 11 
173 ± 14 
P < 0.01 
Difference 
p < 0.01 
p < 0.01 
Values are means ± SEM; difference determined by t-test. 
based on 27 fields photographed both during 
control and hypoxia: 17 from the top and 10 
from the bottom. At this time only one observer 
has made tracings from the fields, so the results 
presented here must be considered preliminary. 
More capillaries are perfused on the bottom 
of the lung than on the top as indicated by grid 
counts: bottom = 153 ± 11 (SEM) versus top 
= 53 ± 7 (p < 0.01). The greater perfusion of 
the bottom of the lung is consistent with the ac- 
cepted concept of a hydrostatic perfusion gra- 
dient favoring blood flow to the dependent 
lung.^ 
With isocapnic hypoxia there is striking cap- 
illary recruitment at the top of the lung: con- 
trol = 53 ± 7, hypoxia = 92 ± 9 (p < 0.01). 
These changes are consistent in that capillary 
counts increase with hypoxia in every case. 
Even at the bottom of the lung where the capil- 
lary bed is more nearly filled to capacity, hy- 
poxia still produces recruitment: control = 153 
± 11, hypoxia = 173 ± 14 (p < 0.01). 
DISCUSSION 
After viewing the films a number of times 
and studying the lung visually, we are con- 
vinced that capillary recruitment with hypoxia 
