MYOCARDIAL IRRITARILITV IIF.GNAUER and COVINO 335 



deviations) of the several coronary A-V electrolyte differences in all gronps prior 

 to immersion. The plus sign indicates higher arterial than coronary venous values, 

 presumably signifying a positive balance with respect to the myocardium. Excep- 

 tions must be made for pH (which if expressed in terms of [H*| would be pre- 

 fixed witli a minus sign) and sodium, the latter ])eing significantly higher in cor- 

 onary venous than arterial blood at normal and low temperatures in all three 

 groups. The magnitude of the sodium difference is such that it cannot reasonably 

 be ascribed to a continuous negative cardiac balance, for the cardiac reserves are 

 not greater than those of other tissues. I'ut an alternati\e explanation for the 

 difference is elusive. 



Since neither the levels (not shown in the tal)les) nor coronary A-\' differences 

 for sodium, chloride and magnesium were demonstrably affected by hypothermia 

 in any of the three groups these ions will not be further considered. The ions 

 apparently affected in one or another group at low temperature were calcium, po- 

 tassium and hydrogen. And these will now be discussed. 



In all three groups the systemic and coronary venous plasma potassium le\x'ls 

 were less at low than at normal temperature by a mean of 0.6 mEq. per liter, in 

 confirmation of other observations."- ^" And since the magnitude and direction was 

 unaffected by pH the decrease must for the present be assumed to be intimately 

 linked with the temperature factor. With respect to coronary A-V differences and 

 systemic blood levels of calcium, a temperature effect was not demonstrable. Left 

 for consideration therefore are coronary A-V differences in hypothermia as influ- 

 enced predominantly by systemic pH. 



In the group-2 dogs (acidotic fibrillators) coronary A-V differences of potas- 

 sium, calcium and hydrogen ions are present. Potassium and hydrogen ion con- 

 centrations are higher in the venous blood and calcium is lower. The potassium 

 difference of 0.2 mEq. per liter is not statistically significant, but becomes so at the 

 5 per cent level if the data of one dog are excluded for analytical purposes. The 

 one exception showed a lozvcr venous than arterial potassium level of a magnitude 

 beyond the range shown by any other dog in any of the three groups. The tremen- 

 dous deviation in the opposite direction accounts also for the large standard devia- 

 tion shown in the table. 



The mean calcium A-Y difference in the same group-2 dogs is 0.4 mEq. per liter 

 and is statistically significant (P=:<0.02). Two dogs in this group failed to show 

 lower total venous than arterial calcium levels. The nonfibrillators. whether acidotic 

 or with normal i)H values, presented either no difference or a slightly higher venous 

 level. 



The pH measurements as given in table II are also revealing. The pH of 

 coronarv venous l)lood at normal temperature for all groujjs. and at 25° for groups 

 1 and 3, averages 0.03 units less than arterial. In no individual instance was a dif- 

 ference as great as 0.05 units observed. Among the fibrillators (group 2) on the 

 other hand are none with a coronary A-V pH difference as low as 0.05 units, and 

 the mean for the group is 0.08. In fact it became apparent that the observer could 

 predict the fate of each dog at 25° from the A-\' pH difference, i.e., whether 

 terminus would be fibrillary or asystolic. 



With the exceptions noted the data suggest that certain acidotic hypothermic 



