EXPERIMENT STATION BULLETINS. 839 



first place, the thermometer cannot be depended upon to read accurately 

 closr than .002° C, and in the second place, less than 1 c. c. of N/25 

 Ca(0H)2 in a large volume of water produces such a small depression 

 that it might not be detected. It is just as accurate, therefore, to add 

 1 c. c. portions and estimate the tenths, using the magnitude of the 

 depression as a guide. 



Finally, the sources of error that might arise from the thermometer 

 and the cooling bath should be mentioned. It has been found that the 

 reading and hence the standardization of the thermometer will suddenly 

 change. This does not happen always, but it does occur occasionally. 

 If the change should occur during a series of titrations of a single soil, 

 and the determination is not checked by a second sample of soil, wrong 

 results will of course be obtained. No definite explanation can be offered 

 for this sudden change of the thermometer. It should be stated, how- 

 ever, that in spite of this occasional uncertain change the Beckmann 

 thermometer is ideally adapted for the determination of the lime require- 

 ment of soils, because its zero point is automatically determined every 

 10 or 15 minutes and for every soil sample. Furthermore, after one is 

 accustomed to the character of the results he can generally detect any 

 change in the thermometer that has taken place. It might be stated in 

 passing that an instrument which is far more accurate and exceedingly 

 more sensitive than the Beckmann thermometer is the thermo-element. A 

 thermoelement can be readily made to read to .0001° C, or ten times 

 more sensitive than the Beckmann thermometer. This instrument, how- 

 ever, is not very convenient to use. 



For the determination of the lime requirement of soils by the freezing 

 point method it has been found essential to keep the temperature of the 

 cooling mixture in both vessels about -2° C. If the temperature of the 

 cooling mixture in vessel F, where the soil is first cooled, is greater 

 than -2° C. the soil will commence to solidify before supercooling has 

 taken place, but if it is only about -2° C. the soil will not start to solidify 

 until it is supercooled to about 1° C, and sometimes it can even be 

 gently stirred while it is supercooling. 



It might appear that the procedure of stirring the soil mass with the 

 thermometer itself and not employing a regular stirrer, might introduce 

 some error. It is believed, however, that no measurable error arises from 

 this source. The use of the thermometer as a stirrer is of great con- 

 venience. 



It should be again emphasized that the soil mass must be thoroughly 

 mixed, care must be exercised that the bulb of the thermometer does not 

 touch the walls of the tube, and that the thermometer should be tapped 

 before the final reading is taken. 



If the above procedure is carefully followed, it is firmly believed that 

 any scientific man can determine the lime requirement of soils v/ith high 

 accuracy and definiteness. 



Having given then a brief exposition of the principles underlying the 

 method, and a rather detailed description of the apparatus and pro- 

 cedure, the presentation of the experimental data obtained, is next in 

 order. 



