COAL-TAR AND WATER-GAS TAR CREOSOTES. 
71 
Table 26. — Comparison of laboratory toxicity tests and service tests on fractions of coal-tar 
creosote. 
Material. 
Boiling point at the still. 
Relative 
position 
against 
Fomes 
annosus. 
Relative 
position 
against 
Xylotrya. 
Relative 
position 
in 
service. 
Below 205° C 
2 
il 
3 
5 
26 
4 
il 
2 
3 and 4 
5 
26 
3 and 4 
26 
205° to 250° C 
5 
250° to 295° C 
4 
295 ° to 320° C 
2 and 3 
Res. above 320° C 
il 
2 and 3 
i The figure 1 indicates the greatest toxicity or service. 
2 The figure 6 indicates the least toxicity or service. 
Table 26 shows that the toxicity and length of service of the 
materials tested are almost completely reversed — that is, the highest 
toxicity gave the least service, and the least toxic substance gave 
the best service. The discrepancy between these two results is, 
however, easily explained. In order to exert its effect upon Xylotrya, 
the toxic principle must be soluble in the body fluids of this organism. 
In a state of purity these toxic principles would have a great and 
immediate toxic effect, but this effect would not be permanent, 
because the toxic principle is taken out by the continual leaching of 
the water. If, however, these toxic principles are mixed with oils 
in which they are more soluble than they are in water, it becomes 
increasingly harder to leach them, because of the retarding action of 
the oil. If, then, the toxic principles of coal-tar creosotes are con- 
sidered to be chiefly the lower-boiling oils, it would be expected that 
in the Petri-dish tests, which are conducted with only a small amount 
of water, the fractions would arrange themselves in the order 1, 2, 
3, 4. 5. But in service tests a greater loss would be expected from 
fraction 1 than from fraction 2, and a greater loss from fraction 3 
than from fraction 4, and so on. The character of the residual oil 
in fraction 1 would tend to approach the character of the oil in 
fraction 5, and in the end the oils would be much the same in compo- 
sition; fraction 1 would contain the least, and fraction 5 the most 
preservative, and the service tests would be expected to show the 
order 5, 4, 3, 2, 1. This was the actual arrangement in the service 
tests. 
The separation of the fractions sharply into those indicated in 
Table 26 is by no means complete. Present knowledge of the action 
of complex mixtures upon distillation indicates that fraction 1 might 
be expected to contain considerable material which should have 
been included in fraction 2, a smaller amount which should have 
been included in fraction 3, and perhaps a small amount of fractions 
4 and 5. Fraction 2 might be expected to contain considerable 
amounts of fractions 1 and 3, and smaller amounts of fractions 4 and 
5. The truth of these statements is illustrated in Table 27, which 
