198 Land Magnetic Observations, 1914-20 



with their 44-calibre Winchester rifles. We drifted a month on the river, stopping only 

 to occupy stations en route, and occasionally for an hour's hunt for monkeys, turkeys, 

 or pigs, which added materially to the variety and quantity of our menu. I had arranged 

 that the Indians furnish their own food, but they adhered to their custom of leaving 

 the responsibility of food supply with the chief of the party and I was obliged to supply 

 almost all their provisions for the voyage. 



We arrived at Riberalta on September 23 intending, if possible, to ascend the Madre 

 de Dios River, but found that all the boats were waiting for the wet season, as that 

 river was at the time almost dry. I therefore continued down the Beni River to Villa 

 Bella, at the mouth of the Mamore River, and from there by train to Guayara Mirim. 

 It was at this point that I suffered the only really serious sickness of the whole trip. 

 Upon partial recovery, I gave up my original plan to ascend the Mamore" to Trinidad, 

 and proceeded down the Madeira River by the Madeira-Mamore Railroad to Porto 

 Velho in Brazil. 



From Porto Velho I went by a good river steamer to Manaos, where I arrived 

 October 29, 1917, and received my first mail since June. After bringing everything 

 up to date, I arranged to ascend the Purus River by taking passage from place to place 

 whenever the unscheduled but frequent steamers came along. The plan was entirely 

 successful, and I went up the Acre River, an affluent of the upper Purus, to Xapuri, a 

 point quite close to the Bolivian border, establishing 8 stations en route. I returned to 

 Manaos and thence proceeded to Obidos and Santarem, reoccupying stations at each 

 place. From Santarem I went up the Tapajoz River for a 10 days' trip by motor-canoe 

 above the head of steam navigation, arriving March 8 at the confluence of the Sao 

 Manoel and Juruena rivers. I made observations here and at 3 other points above 

 Santarem, but was unable to stop at another point as originally planned on account of 

 the very serious illness of one of the men on the boat. The boats on which this trip was 

 made carry about 10 tons of cargo, a crew of 12 or more very skilful boatmen, and from 

 1 to 5 or 6 passengers. In stretches where the current was swiftest, we went close to 

 shore. The gasoline motor working full speed was helped by 12 husky natives, the 

 owner of the boat, and the magnetic observer aboard, all pushing and pulling with 

 poles arranged with various kinds of hooks, prongs, and points adapted for grappling 

 branches, vines, and stones, or for poling. At times a rope was laboriously carried from 

 tree to tree by which the boat was pulled through the rapids. The work was very stren- 

 uous, and on a few occasions a stretch of perhaps 50 feet was passed in no less than a 

 half day, while clothes were literally torn from our backs. 



From Santarem I went to Belem, Para, occupying one station en route. No prac- 

 ticable route for further work in the region of the mouth of the Amazon was open at 

 that time, and after reoccupying the secular-variation station at Pinheiro, and making 

 a short trip by rail to Bragan^a, I proceeded to the Guianas and reoccupied C. I. W. 

 stations at Cayenne, French Guiana; Paramaribo, Dutch Guiana; Georgetown and 

 New Amsterdam, British Guiana. 



At Georgetown I received a cablegram to return to Washington by the safest route. 

 I accordingly left Georgetown June 27, 1918, for St. Thomas, Virgin Islands, and caught 

 a Swedish steamer for New York. I arrived July 8, 1918, and reported at Washington 

 the following day. 



Table 32 (see p. 199) gives a list of the stations occupied, with dates and geographic 

 positions; for magnetic data, see Table of Results. 



The time from sailing from New York to my arrival at New York was over 21 

 months, but a part of this time was spent on my instruction trip (see report by Observer 

 D. M. Wise). The actual field time was 163^ months, which gives an average of about 

 8 days field time per station. During much of the trip rapid travel was impossible, and 



