96 TWENTY-FIVE YEARS OF CHEMICAL INVESTIGATION. 



7. In none of the Ecca waters has any calcium sulphate been founds 

 magnesium carbonate being invariably present, and in most cases also 

 sodium sulphate, often with sodium carbonate. 



8. That the older Karroo rocks yield waters of more varied com- 

 position, calcium sulphate again making its appearance, and even, in 

 some instances, noticeable amounts of calcium chloride. 



9. That the Middle Beaufort beds, on the other hand, show much 

 more uniformly constituted waters. Calcium salts other than the car- 

 bonate are absent, and magnesium sulphate and chloride rare; sodium 

 salts bulk more largely in proportion, but. relatively speaking, the amount 

 of chlorine is less- 



10. That this condition is more accentuated in the Upper Beaufort 

 and Stormberg waters, the proportion of chlorine appearing to diminish 

 as one passes geologically upwards from tlie Lower Beaufort to the 

 Stormberg formation. 



11. That the majority of waters of the Middle and Upper Beaufort 

 and Stormberg series contain sodium carbonate, black alkali, in solution — 

 an important fact in connection wnth agricultural operations in those 

 districts. 



Looked at froni the engineer's point of view, the waters 

 from the Table Mountain series were, without exception, very 

 good ; those from the Mahnesbury beds were fair on the aver- 

 age ; those from the Bokkeveld series either poor or very bad. 

 The Dvvyka waters proved to be, generally speaking, very bad. 

 The Ecca waters were rather poor, those of the Lower Beaufoit 

 beds were bad as a general rule, while those of the Middle Beau- 

 fort beds were good. The Burghersdorp or Upper Beaufort beds 

 supplied water of fair quality, and good water was obtainable 

 from the Stormberg series. The Uitenhage series gave waters 

 of very variable qualitv ; most of them, however, were verv 

 bad. 



In this connection I inay be allowed to repeat just one sen- 

 tence of the address to the Cape Chemical Society in which the 

 above summarised results were first detailed. That sentence bears 

 closely upon some of the opinions animadverted on at an earlier 

 stage of the present paper. Speaking of the numerous water 

 analyses which, when collated, afiforded the opportunity of draw- 

 ing the above conclusions, I said : — 



" We have a number of isolated facts, l:)ut we have to connect them 

 up by piling observation on observation, analysis on analysis, fact on 

 fact ; and this not at haphazard, but systematically, carefully recording 

 all modifying or influencing circumstances ; and we need, too. to educate 

 the public mind with regard to the value and desirability of agricultural 

 chemical investigation." 



ToXICOL(X".Y. 



But I must pass on to another exemplification of exactly this 

 point which the work of the Cape Laboratories has afiforded. In 

 the .'•hort statistical summary of our work above given it was 

 said that 159 poisoning cases had been investigated. Some of 

 these proved to be nugatory, in others arsenic had been the 

 poison administered, in others again strychnine, or some other 

 of the poisons more or less frequentlv met with in toxicological 

 practice. In very many of the cases, however, the problem set the 

 analyst was of quite a dififerent order. In the London courts 

 recently the skill of the analyst who detected a small quantity of 



