CHEMICAI, COMPOSITION' OF RAIN. 1/3 



A further liability to en(M" in the chemical results was fore- 

 seen in connection with the numerous hawks and other large birds 

 which have their habitat on the expansive South African veld, 

 and are apt to use any upward projection (such as a rain gauge) 

 in the treeless Karroo, as a settling place, so fouling the gauge 

 with their excreta. In fact, the Director of the Meteorological 

 Observatory at Johannesburg at first considered it quite impos- 

 sible to keep raiti gauges free from dust and excreta, and believed 

 that samples from even well-watched gauges would give most 

 misleading results. .As far as the dust was concerned, the decision 

 arrived at was to eliminate it by simple filtration, and the bird 

 problem having been experienced in Barbados, and there over- 

 come by the use of large glass funnels, which the birds seemed 

 to refrain from settling on, it was resolved to try the same plan 

 in the South African districts wherever a similar trouble was 

 likely to occur. The appended photograph illustrates the manner 

 in which this difficulty was dealt with at the Johannesburg Ob- 

 servatory. The delivery spout of a five-inch gauge was removed, 

 and the apex of its funnel opened out so as to allow the neck of 

 a seven-inch glass fimnel"^ to pass through. The glass funnel 

 was then firmly held in ]X3sition by three hooks of copper wire 

 passing over the rim. A glass receiver was placed inside the 

 gauge, so that, except for the three copper hooks, the rain came 

 into contact with nothing but clean glass. The gauges were 

 mounted in the usual way on firm stands, with their rims four 

 feet above ground. Two of these gauges were placed a few feet 

 apart, and the rain collected from them was stored in Winchester 

 quarts until the end of eacli month. The glass funnels and re- 

 ceivers were kept clean_ by washing twice a day if no rain had 

 fallen for a day or two, and every efl'ort was made to prevent 

 contamination of the water. Where there was possibility of the 

 gauge being contaminated by birds perching on its rim, the addi- 

 tional safeguard was adopted of surrounding the funnel and 

 gauge by a cylindrical screen of wire netting, the vertical wires 

 of which projected above the highest horizontal strand, in order 

 to prevent the birds from finding any foothold. 



Unfortunately, the untoward circumstances already alluded 

 to came in the way of the analysis of most of the rain v/ater col- 

 lected with so much precaution at Johannesburg; the Ermelo 

 samples were likewise left untouched, and the Potchefstroom 

 samples were only partially examined. But, for future guidance, 

 it is nevertheless useful to record the truncated efforts made to 

 obtain an idea of the composition of uncontaminated water from 

 that centre. 



Instructions were sent to all the collecting centres in tlie 

 Cape, Transvaal and Natal Provinces on the above lines, and, as 

 regards the period of storage, the following memorandum was 



* Preferably one of Jena or resistance glass, so as to diminish the 

 possibility of alkali being dissolved out. 



