290 The National Geographic Magazinb 



reached, and the waters began to slowly 

 recede. This stage of 38 feet was 8 

 feet above the danger line, and within 

 2.6 feet of the great high-water mark 

 of June 27, 1844. 



This flood, while, of course, much 

 more destructive than that of 1844, did 

 not by any means equal it in volume. 

 In 1844 the lowlands were not occupied, 

 there were no busy centers of industry 

 at Armourdale, Argentine, and East St 

 Louis, and consequently the damage 

 done was relatively slight. The rain- 

 fall in 1844, however, was decidedly 

 greater than in May, 1903, and there 

 were no levees along the river to hold 

 the waters. Yet with all bottom lands 

 overflowed, a stage of 37 feet was 

 reached at Kansas City on June 20 and 

 of 41.4 feet at St Louis on June 27. 

 At the same time the Illinois River was 

 from 10 to 15 miles wide from La Salle 

 to its mouth, and from Hardin down 

 united with the Mississippi to form one 

 continuous river. The consequence of 

 an equal amount of rain during the 

 present year and equally well distributed 

 cannot be estimated with any degree of 

 exactness. Nevertheless it is reason- 

 ably certain that several feet would 

 have been added to the stage at St 

 Louis, and that the entire city of East 

 St Louis would have been totally sub- 



merged to a depth of at least 8 or 10 

 feet. 



The annual rise of the Columbia River 

 is always a subject of interest to the 

 people of the north Pacific coast. This 

 rise depends almost entirely upon the 

 melting of the winter snows in the 

 mountains, and there are at times wide 

 divergences of opinion as to the prob- 

 able extent of the rise and its effects 

 upon the Willamette River at Portland, 

 Oregon. In March of the present 3'ear 

 the official in charge of the Weather 

 Bureau office at Portland issued a bul- 

 letin on the subject, in which he stated 

 that from the amount of snow then in 

 the mountains a stage of 24 feet, or 9 

 feet above the danger line, would be 

 reached at Portland about the middle of 

 June. On June 13 the stage was 22.8 

 feet, with the Columbia still rising 

 slowly. 



Against such cataclysms as those at 

 Pacolet and Heppner, flood warnings 

 cannot avail. They are caused by tor- 

 rential downpours upon extremely pre- 

 cipitous watersheds. Millions of tons 

 of water are suddenly poured into a 

 deep reservoir with but a single narrow 

 avenue of escape. The results are then 

 apparent, but they are beyond the prov- 

 ince of human wisdom either to foresee 

 or prevent. 



A SUGGESTED FIELD FOR EXPLORATION 



THE cabled reports tell of the 

 continued activity of Mont 

 Pelee in Martinique, of Colima 

 in Mexico, and of Santa Maria in Gua- 

 temala. For a period now of eighteen 

 months there have been unceasing vol- 

 canic disturbances in a belt extending 

 east and west, from the west coast of 

 Mexico to Martinique, and north and 

 south, from central Mexico to Vene- 

 zuela. Since January 1, 1902, this belt 

 at some point or other along its length 



has been constantly in a state of violent 

 disturbance. The first disturbance oc- 

 curred in January, 1902, when an earth- 

 quake destroyed Chilpancingo in Mex- 

 ico, and caused the loss of thousands of 

 lives. On April 18 Quesaltenango and 

 other towns in Guatemala were likewise 

 ruined and fearful destruction of life 

 resulted. On May 8 occurred the erup- 

 tions of Mont Pelee and La Souffriere, 

 numbering 35,000 victims. In Novem- 

 ber the Santa Maria volcano in Guate- 



