PERIODICITY IN HUMAN BEINGS AND MICE 829 



effects brought about by differences in operating level. With respect to 

 Agren's data, however, the changes shown by relative curves may de- 

 pend upon a few animals, which possibly constitute a subgroup to be 

 ignored, rather than emphasized, i.e., rats with incomplete cortical 

 insufficiency. If, in turn, we should contemplate eliminating those 

 animals with the higher values of liver glycogen from Agren's data — 

 and we would have to do so arbitrarily — the remainder of the values 

 is too low (and the analytical error at such levels is perhaps too high?) 

 for any conclusion other than that, for all practical purposes, glycogen 

 disappears from adrenalectomized rat liver throughout the 24-hr 

 period. Sollberger (1954, 1955a,b), who has substantially contributed 

 to the problem of glycogen rhythm in liver as well as to the statistical 

 aspects of rhythms in general, has recognized this knotty problem 

 (personal communication) which persists despite a vast amount of 

 work on other aspects of hepatic rhythms (Beringer, 1950; Bout well 

 et al., 1948; Dean, 1944a,b; Ekman and Holmgren, 1947, 1949; 

 Forsgren, 1928; Gerritzen, 1940; Higgins et al, 1932, 1933; Holm- 

 gren,"l931, 1941; Holmquist, 1931; Jores, 1940; Mollerstrom, 1940; 

 Petren, 1939; Sjoegren et al, 1938). Whether a 24-hr periodic carbo- 

 hydrate metabolism is essential for cellular rhythmicity, or whether it 

 is merely an adequate carbohydrate supply that (1) is required for 

 such rhythms, and (2) is driven by the adrenal, as are other periodici- 

 ties — this is one of the pertinent questions (cf. Bullough, 1955). In 

 this connection, on one hand, the persistence of glycogen rhythm in 

 the liver of starved animals suggests that feeding per se is not critical 

 with respect to 24-hr periodicity (Agren et al, 1931). On the other 

 hand, in the starved animal, gluconeogenesis from protein seems to de- 

 pend on a periodic corticoid secretion, as do certain other functions 

 that are periodic. Thus, in the light of available knowledge, we cannot 

 go beyond suggesting that an adequate carbohydrate supply is an 

 essential condition for metabolic rhythms in mammalian cells. But at 

 the same time, in the absence of the adrenal cortex, the carbohydrate 

 supply per se need not be a sufficient condition for such cellular 

 rhythms to occur. 



Whatever the intimate mode of adrenal interaction with metabolism 

 may be, our assumption that the adrenal is a pacemaker of a periodic 

 sequence of cellular events may have heuristic value, if the varied ac- 



