BEST 621
working on "pedigreed" patients has yielded fairly consistent
results. Clinicians have detected variations of two or three units
in potency. On the other hand, however, on certain occasions
they have believed that they noticed a variation in the insulin
when no variation existed. The difference was probably due to
some changes in the clinical condition of the patient, of a nature
too subtle to produce signs or symptoms. Nevertheless, the
clinicians are the judges of our laboratory methods of standard-
ization, and when our technique is good enough to produce lots
so uniform in potency that they cannot distinguish between them
there will be little cause for worry.
The Medical Research Council of Great Britain is engaged
in preparing a stable insulin hydrochloride which after stand-
ardization may serve as the official standard.
YIELDS OF INSULTN
There are great variations in the findings of tlose investi-
gators who have reported on the yields of insulin from pancreas.
The variation is due partly to differences in the methods of test-
ing, partly to the varying degrees of purity of the material
under test, and only partly to the amount of insulin extracted
from the gland by the method. It is impossible to evaluate a
method when little or nothing is reported concerning the first
two factors. Dodds and Dickens (18) report a yield of crude
insulin of approximately 5500 clinical units per kilogram of
pancreas. Moloney and Findlay (9) report the demonstration
of 4400 units per kilogram of pancreas. Clough, Allen and
Murlin (19) publish results demonstrating over 4500 units of
crude insulin per kilogram. Dudley and Starling (20) obtain
a maximum yield of about 1500 units of insulin, as finished
hydrochloride, from one kilo of ox pancreas. Acid alcohol as
used in our method extracts from 1500 to 2200 units per kilo-
gram of pancreas. Consistent yields of 1100-1200 units of puri-
fied material (nitrogen 0.006 mgm. per unit) are obtained. In
view of the variation in reported yields of insulin from pancreas,
the matter might be simplified if the yield were always stated
in terms of units of a finished product which has been found
to be completely satisfactory in the clinic.