the attempt natural to withdraw the water after solution and before addition. This, fortunately for the law of Conservation of Mass-Energy, never succeeded, since solution never took place unless the water was eventually added. The question is, of course, instantly raised as to how the thiotimoline can 'know' in advance whether the water will ultimately be added or not. Though this is not properly within our province as physical chemists, much recent material has been published within the last year upon the psychological and philosophical problems thereby posed.
Nevertheless, the chemical difficulties involved rest in the fact that the time of solution varies enormously with the exact mental state of the experimenter. A period of even slight hesitation in adding the water reduces the negative time of solution, not infrequently wiping it out below the limits of detection. To avoid this, a mechanical device has been constructed, the essential design of which has already been reported in a previous communication.6 This device, termed the endochronometer, consists of a cell 2 cubic centimeters in size into which a desired weight of thiotimoline is placed, making certain that a small hollow extension at the bottom of the solution cell - 1 millimeter in internal diameter - is filled. To the cell is attached an automatic pressure micro-pipette containing a specific volume of the solvent concerned. Five seconds after the circuit is closed, this solvent is automatically delivered into the cell containing the thiotimoline. During the time of action, a ray of light is focused upon the small cell-extension described above, and at the instant of solution, the transmission of this light will no longer be impeded by the presence of solid thiotimoline. Both the instant of solution - at which time the transmission of light is recorded by a photoelectric device - and the instant of solvent addition can be determined with an accuracy of better than 0.01 %. If the first value is subtracted from the second, the time of solution (T) can be determined.
The entire process is conducted in a thermostat maintained at 25.00掳 C. - to an accuracy of 0.01 掳 C.
Thiotimoline Purity - The extreme sensitivity of this method highlights the deviations resulting from trifling impurities present in thiotimoline. (Since no method of laboratory synthesis of the substance has been devised, it may be practically obtained only through tedious isolation from its natural source, the bark of the shrub Rosacea Karlsbadensis rugo.) Great efforts were therefore made to purify the material through repeated recrystallizations from conductivity water - twice redistilled in an all-tin apparatus - and through final sublimations. A comparison of the solution times (T) at various stages of the purification process is shown in Table I.
Table I
Averaee 'T'
Purification stage (12 obser- 'T' extremes % error
vations)
As isolated -0.72 -0.25; -1.01 34.1
First recrystallization -0.95 -0.84; -1.09 9.8
Second recrystallization -1.05 -0.99; -1.10 4.0
Third recrystallization -1.11 -1.08; -1.13 1.8
Fourth recrystallization -1.12 -1.10; -1.13 1.7
First resublimation -1.12 -1.11; -1.13 0.9
Second resublimation -1.122 -1.12; -1.13 0.7
It is obvious from Table I that for truly quantitative significance, thiotimoline purified as described must be used. After the second resublimation, for instance, the error involved in an even dozen determinations is less than 0.7%, with the extreme values being -1.119 seconds and -1.126 seconds.
In all experiments described subsequently in this study, thiotimoline so purified has been used.
Time of Solution and Volume of Solvent - As would seem reasonable, experiments have shown that increasing the volume of solvent enables the thiotimoline to dissolve more quickly - i.e., with an increasingly negative time of solution. From Figure 1, however, we can see that this increase in endo-chronic properties levels off rapidly after a volume of solvent of approximately 1.25 ml. This interesting plateau effect has appeared with varying volume of solvent for all varieties of solvents used in these laboratories, just as in all cases the time of solution approaches zero with decreasing volume of solvent.
Time of Solution and Concentration of a Given Ion - In Figure 2, the results are given of the effect of the time of solution (T) of varying the volume of solvent, where the solvent consists of varying concentrations of sodium chloride solution. It can be seen that, although in each case the volume at which this plateau is reached differs markedly with the concentration, the heights of the plateau are constant (i.e. -1.13). The volume at which it