BRITISH GUIANA AND BRAZIL — SMITH 337 



season. Field work stopped in May, when survey parties returned to 

 Georgetown for 10 weeks, leaving fresh boats' crews to provision all 

 camps during the high water. Base camp was moved to Oronoque. 

 During the recess the British and Dutch sections placed pillars at 

 the mouth of the Courantyne to mark the direction of the boundary 

 through territorial waters. In August 1936 the survey parties again 

 took the field, but with a very different outlook. All ration depots were 

 well stocked; Indian and Negro labor was more plentiful and better 

 understood by the officers; and much of the upper river was cleared 

 for canoe communication. The international watershed was located 

 at the source of the Oronoque by the beginning of September, and an 

 astronomical pillar erected. The boundary was traced eastward, and 

 rapid progress made through a country of small lumpy hills 200 to 

 300 feet high, separated from one another by creeks and swamps. 



Bush survey has a rough-and-ready technique of its own. A visi- 

 bility of 20 yards, often less, is not conducive to rapid topography, 

 and forcing a passage through the tangled undergrowth is so slow 

 and fatiguing that the extent of the day's excursion is very limited. 

 Nevertheless, it is surprising what good results can be obtained with 

 care and a simple routine. The boundary was to be marked by 

 buried stone on the international watershed at not less than 5-mile 

 intervals, the position of such marks to be indicated by a small con- 

 crete pillar. Every fifth mark or 25 to 30 miles of boundary was to 

 be fixed in position by astronomical observations and the whole con- 

 nected by an instrumental traverse to the order of 1/250. In addi- 

 tion it was the practice of the British section to map the topography 

 for 1 to 11/^ miles on either side of the watershed, but during the last 

 season this was discontinued. 



The biggest problem was to find the boundary. Running water 

 was plentiful, but to establish whether it was British or Brazilian 

 without tracing its course for some miles was impossible. The creek 

 heads of each country were twisted and contorted into a maze of 

 steep little hills, saddles, and swamps, and every hill was, of course, 

 a watershed. A creek would often wind its way in a general south- 

 erly direction for a mile or more, only to turn and prove itself to 

 be British. One well-remembered creek actually split in two, the 

 main branch flowing north to the Atlantic 400 miles away, while a 

 smaller branch bubbled under some stones and commenced its 2,000- 

 mile journey to Para at the Amazon's mouth. In desperation the 

 boundary was placed up the center of its bed. 



Boundary location was done by Indians, and it was a very arduous 

 task, as it entailed cutting through virgin bush from morning to 

 night, finding creek heads and tracing the water till it declared its 



