2 Mr Bar croft, Apparatus for Analysis of the Gases, etc. 
or to put the matter in another way, my object has been to secure 
the same error in the figures when worked out as percentages 
as others have done, while using only a fifth of the blood for 
analysis. 
The most obvious source of error is, of course, the contamina- 
tion of the blood gases with small quantities of air, and it will 
be apparent that, in an apparatus the parts of which are de- 
tachable, it is almost impossible to avoid small leakages. 
In the present apparatus there are no joints to be made, as 
the whole is in one piece. This arrangement would be simple 
were it only necessary to collect one sample of blood ; the present 
apparatus however provides for the collection of nine samples, 
six from a vein, and three for comparison from an artery. 
With regard to the errors of the apparatus, the greatest is that 
of measuring the blood, as the surface of the mercury beneath 
it is sometimes broken by films of blood ; the extreme error 
here is '05 c.c. Another error is produced by the fact that a film 
of blood is left over the glass in a tube leading to the chambers 
and in the burette. This is a very variable quantity and depends 
in the first place upon the individual apparatus, and in the second 
place upon the tendency of the blood to coagulate. If the blood 
is very fluid, and the apparatus carefully made with the tube 
straight and of 2 millimetres diameter, the amount of blood then 
left as a film is about '05 c.c. Each sample will therefore be con- 
taminated to that extent by the previous one. 
In regard to the collective error of the whole apparatus, the 
following analysis of defibrinated blood will serve as a guide. 
Analysis of three 
samples 
of the 
same defibrinated blood 
co 2 
23*7 
23-9 
24*0 c.c. 
0 
20*4 
202 
20 6 c.c. 
N 
1-9 
1-7 
1*9 c.c. 
On another occasion three 
samp! 
es gave respectively, 
C0 2 
52-6 
52 6 
527 c.c. 
0 
11-9 
11 '9 
11*9 c.c. 
N 
1-6 
24 
T9 c.c. 
Description of Apparatus. The Pump. 
The Toepler Pump has been employed, without any modifica- 
tion except that the drying apparatus is so arranged that the 
sulphuric acid can be renewed without introducing more air into 
the apparatus than what is dissolved in the acid. 
A three-way tap, Fig. 1 (c), leads to a water-pump which 
gives a vacuum, almost up to the tension of aqueous vapour. 
