192 CALCULATION OF DATA 



data are sometimes necessary. One type is the transformation of dimen- 

 sional units described in the preceding section. Another type is a conver- 

 sion of the data produced by an instrument into other terms. For example, 

 a stripchart recorder might record millivolts when we really want to 

 know a change in chemical composition. In the properly designed instru- 

 ment, the actual output will be related to the information being sought 

 by some unvarying mathematical relationship or "transfer function." In 

 the ideal instrument, the output is directly proportional to the response, 

 the transfer function is linear, and the only correction required is a multi- 

 plication by a constant. 



We must put up with random errors or variations of measurement. 

 We can minimize such errors or estimate the size of the variations, but 

 we cannot eliminate this source of error and we cannot correct for it. 

 A systematic error, however, leads to an inaccuracy of measurement 

 which results from some defect in the standards used for comparison. 

 The too-short meter stick was used as an example in Chapter 4. We can 

 estimate the magnitude of errors of this sort and then correct for them. 

 If we know, for example, how much too short our meter stick is, a 

 simple computation gives us corrected values. 



The following extensive example illustrates several kinds of manipula- 

 tions of data. A set of measurements of rate of photosynthesis was made 

 using the manometric method (Chapter 10). The data were recorded 

 on a printed form (one suggested by Umbreit, Burris, and Stauffer 0» 

 and some of the computations were performed on the same sheet. Figure 

 14-1 shows such a record sheet in an abbreviated form. The thermo- 

 barometer vessel contained only water. The other manometer contained 

 Chlorella cells suspended in KHCO3 solution so that the only gas ex- 

 change affecting the pressure was the production of oxygen. The con- 

 stant, K().^ = 1.32 lA 02/mm of manometer Huid, is a transfer function. 

 The "raw data" include the time, the pressure on the thermobarometer, 

 and the pressure on the experimental manometer. The second column 

 under the thermobarometer and also under the experimental manometer 

 is the amount of change since the beginning of the experiment. In the 

 third column under the experimental manometer the first correction is 

 made. The pressure change exhibited by the thermobarometer represents 

 a systematic error, that is, a change in room air pressure since the begin- 



1 W. W. Umbreit, R. H. Burris, and J. F. Stauffer, Manometric Techniques, 

 3rd ed. (Minneapolis: Burgess Publishing Company, 1957). 



