220 Mr. Alexander Siemens ' [May 15, 



76° 10' W. long., although the Ucayale, where it unites with the 

 Maraiion at Nauta (4° S. lat., 73° W. long.), is quite as important as 

 the Marafion. If the greatest distance from the mouth is to decide 

 the question, then the source of the Apurimac, an affluent of the 

 Ucayale, can lay claim to being the origin of the Amazon, rising in 

 Peru in 16° S. lat. and 72° W. long 



From the Lake Lauricocha the main direction of the Tunguragua 

 and the Maranon is to the N.N.W., until the river turns eastward, 

 and shortly after passing Jaen breaks through the Andes, entering 

 the plains of the Amazon valley by the Falls of Manseriche, a short 

 distance west of Borja. Its further course is a little north of east, 

 until it pours its yellow waters into the Atlantic under the equator 

 between the Cabo do Norte and the Cabo Maguari, which are 158 

 miles apart. This distance is just about equal to the distance from 

 Land's End to Cape Clear in Ireland, or from Brighton to Falmouth. 

 Even west of the island of Caviana, which lies in the mouth of the 

 river, together with the island of Mexiana and several smaller ones, 

 the width of the main stream is over 50 miles, equal to the distance 

 from Portland Bill to the Cap de la Hague. The part of the Amazon 

 flowing north of the Island of Marajo may therefore be compared in 

 width to the Channel, but in depth and volume of water it far sur- 

 passes it. It is a disputed question whether the water flowing south 

 of Marajo, commonly called the Para river, should be considered as 

 part of the Amazon or not. A network of natural canals, "the 

 narrows," connects the two waterways west of Marajo, but the 

 influence of the tide makes it difficult to decide whether part of the 

 water of the Amazon finds its way south of Marajo or not. Along 

 the old course of the Amazon, commencing at the foot of the Andes, 

 a similar network of islands and canals is formed on both sides of the 

 river, as the whole country is almost level, and is consequently 

 inundated during the rainy season for hundreds of miles by the rivers 

 flowing through it. The most notable exception to this general state 

 of things occurs at Obidos, where the whole volume of water is com- 

 pressed into one channel a little over a mile wide, and said to be 

 about forty fathoms in average depth. 



A sounding taken opposite Obidos, about a third of the distance 

 across the river, showed a depth of 58 fathoms, measured by a steel 

 wire and Sir William Thomson's sounding machine. As the 

 current of the river averages three knots in the main channel, it is 

 not easy to take soundings by an ordinary lead line, and even with 

 the steel wire an extra heavy weight (33 lbs.) has to be employed, or 

 the results are not reliable. Besides the wire sounding machine, 

 James's Submarine Sentinel was used on the preliminary voyage, 

 wherever serious doubts existed about a channel through which the 

 cable was to be laid. Usually the sentinel was set at five fathoms, 

 and when it struck a bar the ship was stopped, and a series of soundings 

 taken to ascertain the exact depth of water and the extent of the 



