WATER REQUIREMENT AND ADAPTATION IN EQUISETUM 11 



the soil when heated by the sun's rays. The plants themselves 

 are very sensitive to heat so that the wax cannot be applied near 

 the stems until it has cooled sufficiently to form a scum over the 

 surface. The waxed paper also proved a satisfactory substitute 

 for rubber on the Ganong shells. 



The plants were grown in loam which had been sifted through 

 a two-millimeter screen in order that absorption by the roots 

 would be as nearly uniform as possible. Records of humidity 

 and temperature during each experiment were made by a Draper 

 self-recording hygrometer and thermometer. 



The transpiration of two or more plants was compared on the 

 basis of the number of grams of water lost per unit area. The 

 stem and branches of Equisetum are almost cylindrical so that 

 the determination of the area of the surface was comparatively 

 simple. The diameter of a stem was measured with a vernier 

 micrometer at the middle of the lowest and topmost internodes. 

 These represented the extremes in size, and from the average, 

 the average circumference of the stem was calculated. This 

 multiplied by the length of the stem gave the area of the 

 surface. 



Leaf areas were found in two ways, by the planimeter and by 

 weighing paper prints. Leaves with comparatively simple out- 

 lines were traced on paper and these retraced with the planim- 

 eter. After practice this could be done readily and the average 

 reading of three tracings was recorded. This method has the 

 advantage of giving the area directly but also has disadvantages 

 in that the square inches must be converted into square deci- 

 meters, and that it is difficult to trace leaves with very toothed 

 margins. In the paper weighing method either tracings or blue 

 prints of the leaves were made and cut out with a scissors. The 

 weight of these prints divided by the weight of one square 

 decimeter of the same paper, multiplied by two gave the trans- 

 piring areas of the leaves. Table 1 shows how the two methods 

 check with one another. 



This comparison of the two methods shows an average differ- 

 ence of only three hundredths of a square decimeter. This was 



