162 Plant Tissue Culture 



distort and enlarge the apparent image. By this 

 method, measurements can be made with sur- 

 prising accuracy. The accuracy of measurement 

 has been determined in two ways. In the first 

 place, a series of 10 roots varying in length from 

 18 to 110 mm. was measured. The flasks were 

 shuffled so that the order of measurements was 

 changed. The roots were remeasured and the 

 process repeated seven times. The measurements 

 were then subjected to statistical analysis. The 

 results are shown in Table 2. They indicate that 

 measurements made by an observer on a single 

 set of roots will vary from measurement to mea- 

 surement by somewhat less than 2 per cent but 

 give no indication of the deviation from the true 

 length. 



In the second place, a series of 20 cultures was 

 measured in the routine way by two different ob- 

 servers and the results recorded. The roots were 

 then removed from the nutrient, placed in a vessel 

 of water, and stretched out along a ruler in such 

 a way that an accurate measure of the true length 

 could be obtained. These last measurements were 

 then compared with the measurements of the two 

 observers made by the standard method, giving a 

 basis for expressing the average error of the 

 method in the hands of these observers. (Ob- 

 server A was an experienced worker; observer B 



