32 THE ROYAL SOCIETY OF CANADA 



Some time after the injection of insulin the rabbits often show- 

 characteristic symptoms. A preliminary period of hyperexcitability 

 gives place to a comatose condition in which the animal lies on 

 its side, breathing rapidly (often periodically) with sluggish con- 

 junctival reflex and widely dilated pupils (rectal temperature normal) . 

 On the slightest stimulation, as shaking of the floor, violent clonic 

 convulsions supervene in which the animal either throws itself over 

 and over, or lies on its side with head markedly retracted and the 

 limbs moving rapidly as in running. These convulsive seizures 

 usually last 1-2 minutes, and they often come on without any apparent 

 stimulation, when the interval between them is about 15 minutes. 

 They frequently terminate in death from respiratory failure. 



Out of a total of 123 rabbits receiving insulin convulsions were 

 observed in 26 cases, and the maximum percentage of blood sugar at 

 which they occurred was 0.047 (except in one animal in which 0.067 

 was found), and the minimum percentage at which there were no 

 convulsions was 0.037. 



This close parallelism between the percentage of blood sugar 

 and the incidence of convulsions suggests that a causal relationship 

 exists between the two. This view is supported by the observations 

 of F. C. Mann on dogs rendered hypoglycaemic by removal of the 

 liver from the circulation (Proc. Amer. Physiol. Soc. Dec, 1920), and 

 by the fact that we have found that subcutaneous injection of dextrose 

 (4 gms. in 20 per cent, solution) restores the animal to the normal 

 condition within a few minutes of the injection. Occasionally 

 recovery may ensue without injections of dextrose, but this is rare. 

 The animals restored by dextrose may subsequently relapse into 

 convulsions which can again be removed by dextrose. Injections of 

 saline solutions or of pentose sugars have no efïect. 

 2. The Effect of Insulin on Hyperglycaemic Animals. 



For these experiments the rabbits were fed on oats and sugar so 

 as to ensure an abundant accumulation of glycogen in the liver. 

 Portions of liver were also removed after death for determination of 

 glycogen. 



The methods employed to cause hyperglycaemia were asphyxia, 

 carbon monoxide poisoning, injection of epinephrine (adrenalin) 

 piqûre and ether. Having satisfied ourselves, by at least four observa- 

 tions in each group, that the above methods cause marked hyper- 

 glycaemia in untreated rabbits we then repeated them on rabbits pre- 

 viously injected with insulin. In every case we found in the injected 

 animals either that there was no rise whatsoever in the percentage of 



