850 PERIODIC FUNCTIONS IN MAMMALS 



carried out since each point represents one determination made on a 

 pool of serum. 



With particular reference to Fig. 25, it is important that the day- 

 night difference in mitotic count of liver parenchyma had changed 

 sign and was statistically significant at 8 days after inversion of light- 

 ing, while in the same animals mitoses in pinnal epidermis were as 

 yet not significantly reversed. In other experiments, at 23 days after 

 lighting inversion (Halberg, Bittner, and Smith, 1957), and at 12 

 days thereafter, the shifted timing of mitoses in pinnal epidermis was 

 clearly apparent. That a shift of mammalian 24-hr rhythms occurs 

 only gradually (rather than abruptly) after the sudden change in 

 lighting regimen has previously been shown elsewhere for rhythms at 

 the level of the body as a whole (Calhoun, 1945; Johnson, 1939) and 

 in our laboratory, with special reference to cellular rhythms (Fig. 25) 

 (Halberg, Bittner, and Smith, 1957; Halberg, Visscher, and Bittner, 

 1953). Moreover, such shifts of rhythm, by shifts of lighting schedule, 

 are in the mouse the rule rather than the exception as regards 24-hr 

 periodic phenomena at different levels of organization. Other things 

 being comparable, we note, as was predicted, that the shift times of 

 various rhythms need not be the same, even if we are dealing with the 

 same phenomenon, mitosis, in the same inbred stock. If, as is likely, 

 this assumption is correct, it follows that subtle differences among the 

 factors governing mitotic rhythms in various tissues are amenable to 

 analysis, inter alia, by explorations of their shift times. 



Hastings and Sweeney (1957b) have shown that in the unicellular 

 Gonyaulax certain 24-hr rhythms (which persist in constant dim light) 

 can be shifted in their timing by a shift in the regimen of alternating 

 light and darkness. With the lighting conditions used, in Gonyaulax 

 the shift is accomphshed in 1 day. Brown and Webb (1949) had 

 shown that the same applies to 24-hr periodicity in crabs but with a 

 slightly longer shift time; the shift in crabs takes a few days, and the 

 shift time is a function of illumination. Corresponding data by John- 

 son and others (Calhoun, 1945; Johnson, 1939) demonstrated that 

 the activity rhythm of rodents also may be shifted, but again, more 

 slowly. Under our conditions, the shift in activity rhythm is completed 

 within 6 to 7 days. With the data of Figs. 19-25, viewed against the 



