278 VETEEINAEY TOXICOLOGY 



apparatus in this respect is best guaranteed by the per- 

 formance of a blank test. If this gives a negative result — 

 e.g. for arsenic — the reagents and apparatus are good. A 

 blank ought to be carried through from time to time as a 

 matter of routine precaution. Fortunately, to-day the manu- 

 facture of high grade chemicals has reached such a pitch 

 of excellence that it is rarely necessary to specially purify 

 one's reagents. 



Qualitative Analysis. — This aims at the detection of the 

 presence of a poison. It is the fundamental and really 

 difficult task of toxicology. Quantitative analysis follows 

 the qualitative, and its chief value is to guide the expert in 

 the formation of an opinion as to the significance of the 

 qualitative revelations. It has also a subsidiary value — 

 that of providing a figure to be produced in evidence. The 

 position of a witness who states that from 8 ounces of a 

 material he separated jV grain of strychnine is stronger 

 than that of the witness who states that he found a 

 ' distinct trace ' of that agent. No wise man would commit 

 himself to the statement that the material contains a given 

 weight of the poison ; he will name the quantity he actually 

 separated, and he will also be well advised to state that the 

 weight given is his estimate, and be prepared to state how 

 he determined it, and what is the probable error of his 

 determination. 



All quantitative data in toxicology are to be regarded as 

 approximations. But even so they are indispensable and 

 valuable approximations, for the reasons set forth above. 

 The quantitative methods available for the estimation of 

 traces of material are — 



1. Direct weighing. This is the most satisfactory and 

 sometimes the only possible method. It may be used for 

 the measurement of quantities of lead and many other 

 common metals. A residue of an alkaloid may also be 

 weighed, but the weight ought not to be regarded as very 

 exact, partly by reason of the normal error in weighing, 

 but chiefly because the residue is rarely pure. Alkaloid 

 extracts always contain traces of basic substance of animal 



