REPORT OF THE SECRETARY 



105 



method whereby the correction could be computed from our daily 

 solar constant observation themselves. This has been done. 



Figure 1 shows a portion of the solar energy curve observed at 

 airmass 2.0. The ozone absorption occurs between places 20 and 20 

 of this curve, but is barely, if at all, visually discernible thereon, 

 even when ozone is most prevalent. However, its etl'ects can be 

 made both discernible and measurable by the following simple pro- 

 cedure. If we take half a dozen of the best days observed in autumn, 

 when the ozone is near its minimum amount, and compute the mean 

 values of the heights of the energy curve in the blue, green, yellow, 

 and red we obtain thereby standard values proportional to the dis- 



FlouKc 1. — Solar energy curve In the region of absorption by atmospheric ozone 



tribution of energy in this region. Those standard values, as thus 

 extended from the violet of the spectrum to the red, overlap at 

 each end the ozone region. Next consider the observations of the 

 heights of the energy curve at these selected places on any given day 

 of observation. We divide them by the standard values just referred 

 to and the result is a series of ratios, near unity, but tending some- 

 times to be lower in the violet than in the red, or vice versa. If 

 plotted against the spectrum place-numbers, these ratios may lie 

 nearly in straight lines. But if the ozone content of the atmosphere 

 on the day examined is different, being larger or smaller, than that 

 of the average of the standard days, then the ratio plot just described 

 presents a loop below or above that straight line which is fixed by 



