I have to admit that there are conditions not yet known, which sometimes diminish the sensitivity of the mice. We have recently introduced the improvement that in each assay a number of mice from the same box are injected with suitable doses of a standard preparation which will bring out any general change in sensitivity. This will I trust lead to the clearing out of the difficulties. Generally however we are very satisfied with the method and get a good agreement with the clinical estimates of strength.
Our large scale production of insulin is getting on fairly well. We are now providing about 20000 physiological units per week for clinical purposes and the quality is increasing at a rapid rate. Our yields are at present not as good as we could desire.