1920] SHULL— SEEDS 375 



compu 



from 





data fall more and more below the computed values. This 

 falling off of the actual intake marks the beginning of the effects 

 of approaching saturation. It is evident that tangents to the curve 

 may safely be computed up to about 35 per cent of intake, but 

 beyond that point the tangents could not be used for comparisons 

 of the rate of intake in different curves. 







For the absorption at 20 C. the substituted values for the con- 

 stants make the equation read y = 61 . 5 log I0 (o. 0136X+ 1) -f 1 . 46, and 

 the corresponding equation f or 3 5 C . is y = 74 . 5 log™ (o . o 1 84a- + 1 ) + 

 2.25. The closeness of the computed intake to the data of observa- 

 tion in each case is shown in table VIII. 



In the 20 curve the effects of approaching saturation first 

 manifest themselves at about 37.5 per cent, and in the 35 curve 

 at about 40 per cent of intake. In each curve the computed 

 values are strikingly close to the actual data. The uniformity of 

 absorption and the agreement of the calculated intake to that 

 observed has been a surprising feature of the work; and since the 

 final break due to approaching saturation is always at or beyond 

 35 per cent, I have felt confident of accuracy in measuring tangents 

 of the curves to that point. 



In the later work the data could not be so satisfactorily repre- 

 sented by means of a single equation. By the use of two or three 

 successive equations, however, each joined to its successor in a 

 point of equal tangency, a very close agreement between calculated 

 intake and experimental data was obtained. For the purpose of 

 calculating tangents, and rates of intake, this composite curve 

 is just as satisfactory as if it were developed from a single equation. 



The 5 curve will be considered first. The three empirical 

 equations used are as follows: 



(1) y=i43 log™ (0.078x4-1) + 1. 398 



(2) y = 35-°7 logic (o.oi2is+i)+4.i95 





(3) ^ = 87.95 logxo (o. 0023*+ 1) +8. 625 



The first two curves have equal tangents for # = 35.35, and 

 the last two for #=150.80 (minutes). The breaks in the curve 



