2 FRANCIS E. LLOYD 



attempt at explanation would have involved experimentation ad 

 hoc, which was impossible at the time. As the ocotillo (Fouquieria 

 splendens) was the subject of study, no further opportunity was 

 presented for additional investigation till the summer of 1910, 

 when, at the courtesy of the Carnegie Institution of Washington, 

 I was able to extend my study of this plant at the Desert Botanical 

 Laboratory. The purpose to determine the ratio, hour by hour, 

 between the volumetric and gravimetric readings was then carried 

 out. 



In making these and the synchronous instrumental readings 

 I had the assistance of Dr. W. H. Brown. The drawing for the 

 accompanying figure was kindly prepared by my colleague Mr. 

 C. G. Gaum. To these gentlemen I tender my thanks. 



In addition to these readings made for three pieces, material 

 for studying the condition of the stomata was taken at suitable 

 intervals, pieces of epidermis being fixed by absolute alcohol, 

 the method previously used. (Lloyd, I. c.) The whole series of 

 readings extended for the space of twenty-seven hours, beginning 

 at 7 p.m., August 21. The data presented in the tables are those 

 for the twenty-four hours beginning at 10 p.m. Three pieces of 

 ocotillo, each about 50 cm. long, were carefully selected from good 

 plants which were evidently well supplied with soil moisture, as 

 they were in full leaf, well colored, and showed no signs of yellow- 

 ing. A slight change in color from green towards yellow indicates 

 that the plant is beginning td respond to a lessening water supply. 

 The selected pieces were set up in burette potometers, as described 

 in my earlier paper, and, after noting that the changes in the rates 

 in the three pieces were approximately parallel, they were regarded 

 as ready for use. All the readings were made in the same rotation 

 each hour. When all the readings had been recorded, each piece 

 with its attached burette was carefully weighed. Immediately 

 at the close of each weighing a second reading was recorded, to 

 serve as a check on the first reading and for close comparison with 

 the gravimetric data. This was of course necessary, since the 

 intervals between the hourly volumetric readings and those ob- 

 tained by weighing were not quite equal. The pieces were set 

 out in the open, and remained there during the whole period of 



