It will be evident that in this experiment, purposely conducted in
AH exaggerated form, the difference produced is very small, and
therefore it may safely he concluded that no appreciable effect
would be caused by a very thin film of paraffin in experiments
M'hich lasted for 3 hours on an average.
III. Glass of Thermometer introduced into the Acid. For the greater period of time during which the experiments were conducted the thermometer hole was closed with a quartz plate; as the thermo- meter, made of lead glass, was introduced for only a short time it may reasonably be assumed that the quantity of glass dissolved out under these conditions would be insignificant.
IV. Personal and Experimental Error due to Micrometer Screw. The majority of adjustments were probably accurate to 2" or a possible extreme error of 4", a magnitude which would change the value of p for the weakest acid by 0-000019 and for the strongest of 0-000008 ; the values for \L are therefore correct in most cases to one unit in the fifth place with a variation in the sixth place.
XI. Temperature Coefficients.
In order to determine the above values, the sample alluded to above may be given as an illustration and the following results cited :
| Temperature (corr.). | Value of fj. D . |
| 11 -G | 1-407157 |
| 18 -.1 | 1-409349 |
| Differences... | 0-002808 |
from which a decrease in \j. of 0-000432 is obtained for increase of 1 temperature; two other determinations gave ^ D 15 ' 3 = 1*405558 and /x D 13 = 1-405567. Accepting this temperature coefficient there are thus four values for the reduction of f* D to the standard tempera- ture taken, 14*2, and the degree of concordance is shown below :
| Temperature. | MD reduced to 14'2. |
| 11-6 | 1-406034 |
| 13-0 | |
| 15-3 | |
| 18-1 | |
| Mean ... | 1-406037 |
In like manner the temperature coefficients expressed in terms of /*;> were determined for all the other samples of acids and the values reduced to the standard temperature as shown above. The following table gives these coefficients :
