94 



NATURE 



[_Nov. 25, 1880 



been weathered oat. In some places a third line runs alongside 

 the two, but this is much less distinct and persistent The 

 double band resembles nothing more nearly than the hollow 

 impression that would be left by double bars of iron placed 

 closely together and neatly inserted in the rock for clasping some 

 structure "on it, if the iron were subsequently removed ; or, as 

 you suggest, tlie marks of a gouge driven by a carpenter across a 

 board. "The bands, when looked narrowly into, consist of very 

 fine close hair-like lines, continuous and parallel to their sides, 

 resembling very minute stria; left by glaciation, and look as if 

 caused by some object drawn along the original red sand, before 

 it became the present indurated rock. 



"A similar double line runs parallel to this one, about tv.o 

 feet lower down, seven feet long, and a third parallel double 

 line on the other or upper side, three feet long, both of the 

 same breadth as the first. Besides those pointed out by you, 

 which occur on the same flat of sandstone, other Imes exist 

 farther dDwn, on the other side of the pool below this rocky 

 flit, on a similar bed of sandstone, part of the same layer— one 

 three feet in length, another six feet, running more or less 

 parallel to those above. Indications of others may also be seen, 

 and, no doubt, several more may be discovered on more careful 

 examination. 



"What they are I can scarcely even surmise, having seen 

 nothing of the same kind elsewhere. They do suggest the 

 possibility of their being the indentations of the caudal append- 

 age of some huge creature, similar to the hallow tail-lines between 

 the footprints on the sandstone at Tarbatness and along the shores 

 of Morayshire — a suggestion strengtheued by the fact of the 

 existence, on both sides of the line, of numerous rounded 

 hollow marks, very like the footprints on these reptiliferous 

 rocks, occurring, as in them, at intervals. But the continuous 

 even breadth and square section of the bands would seem to 

 render this impossible. Then they might be the depressions left 

 on the soft sand by the hinder portions of the shell of some large 

 crustacean — a more likely cause, rendered more probable by the 

 existence of very good ripple-marks on the same sandstone, in 

 the same and neighbouring layers. The stria:-like lines of which 

 the grooves consist would seem to point to some moving agent, 

 organic or physical. They may, however, be the casts or 

 impressions of some great land reed or sea fucoid, the hairdines 

 being the marks of the fine structure of its stem or the parallel 

 veins' of its leaves. It would be desirable to have the super- 

 incumbent layer of rock carefully removed where the bands in 

 question disappear under the upper rock, which might shed some 

 light on the nature of the strange marks. I was sorry I could 

 not spend more time on their examination." 



The impressions occur about 300 or 400 yards above the 

 Victoria Falls, and immediately beside the last of three lesser 

 waterfalls on the west side of the stream. 



THE QUANTITIES OF WATER IN GERMAN 

 RIVERS 



AN attempt has recently been made by Herr Graeve (Da- 

 Civil- Ingenieur, 1S79, p. 591) to determine the amount of 

 water in German rivers and its apportionment in different 

 seasons, a question very important for navigation, and also of 

 much scientific interest. His research comprehends the chief 

 rivers of Germany, excluding the Danube, which begins to be 

 navigable only outside of Germany, and including the Vistula 

 and the Memel. He first calculated, from the mean heights of 

 water, the quantities of water flowing out per second, and he 

 adds a table in which the amount of outflow is shown in relation 

 to the extent of the corresponding river territory. When the 

 amoinit of outflow per loo sq. km. of the region of precipitation 

 is calculated the following values are obtained: — (i) the Rhine 

 at Coblentz above the Moselle mouth delivers per 100 sq. km. of 

 land I 070 cub. m. of water in a second ; (2) the Weser at 

 Mmden, oS26cub m. ; (3) the Elbe at Sorgau, 0-579; (4) the 

 Elbe at Barby, 0-554; (5) the Oder at Steinau, 0-460 ; (6) the 

 Oder below the Warta mouth, 0-413 ; (7) the Warta near its 

 mouth, 0-344 ; (8) the Vistula at Montau ijpitz, 0-538 ; (9) the 

 Memel at Tilsit, 0-600. 



From these numbers it appears (a) that the average outflow of 

 different rivers, from equal portions of theu- territory, differs 

 much more than is usually thought, for in the Middle Rhine it 

 is about three times, in the Middle Weser two and a half times, 

 and in the Middle Elbe, as also in the Lower Vistula and Memel, 

 more than one and a half times as much as in the Lower Warta. 



On the whole, it decreases from the Rhine to tlie Warta, and 

 from the latter increases again to the Memel. (/') In one and 

 the same river the quantity from equal portions of land seems as 

 a rule to decrease down stream, (c) All calculations of quantity 

 of outflow in streams, based merely on extent of the region of 

 precipitation, must as a rale give incoirect results. 



It was important to try and determine the relations of the 

 quantity of outflow to the rainfall of the con-esponding regions, 

 and Herr Graeve, doing so by a method which he describes, 

 obtained the following percentage numbers, corresponding to the 

 above series of rivers : — (i) = 38-5 per cent. ; (2) = 37 p. c. ; 

 (3) = 30 P-c. ; (4) = 28-5 p.c. ; (5) = 27-2p.c. ; (6) = 2i-4p.c.; 

 (7) = 21 p.c. ; (8) = 29 p.c. ; (9) = 32-5 p.c. 



From this the following conclusions (briefly) are drawn : — 

 {a) The percentage proportion of the amount of outflow to 

 the rainfall differs very considerably in these several rivers, 

 though far less than the amount of outflow from equally large 

 regions of these rivers ; hence the differences of the laller can 

 be due only in part to differences in the rainfall. 



(/<) The percentage decreases from the Rhine to the Warta, 

 and increases again from the latter to the Memel. In one and 

 the same river a decrease is perceptible down the stream, at 

 least so far as the phenomena in the Oder and the Elbe are 

 general. 



((-) Since in a mountainous region a greater part of the atmo- 

 spheric precipitates is carried off by rivers than in the plain, the 

 steady decrease in the percentage proportion of outflow to rain- 

 fall in the direction from the Rhine to the Warta must be 

 primarily attributed to the increasing flatness of the region ; so 

 too must the decrease of the percentage down stream. The 

 influence of more or less wood on the land could not be precisely 

 determined. 



(d) The marked increase of the percentage in the direction 

 from the Warta to the Memel cannot be explained by the 

 orographic conditions of the region of precipitation, because 

 this region in the case of the Memel is not at all hilly, 

 and in that of the Vistula only a little more hilly than 

 that of the Warta, but since the amount of the evaporated 

 part of atmospheric precipitates is considerably influenced by 

 the mean temperature of the region- of precipitation, and this 

 in the region of the Vistula and the Memel is lower than in that 

 of the Warta, the increase of percentage in question from the 

 Warta to the Memel must mainly be attributed to climatic 

 conditions. 



(;•) While the percentage in <)uestion must be chiefly governed 

 by orographic and climatic conditions, there can be no doubt 

 that other factors also act, <-.^., the relative amount of moisture 

 in the air, which influences the degree of evaporation, and in 

 general must decrease from the rainy Rhine region to the dry 

 region of the Warta ; further, the amount of plantation, which 

 in°the regions of the Vistula and Memel is larger than in those 

 of all other German rivers ; lastly, the nature of the ground, 

 allowing more or less passage to the precipitates ; the influence 

 of all these factors, however, cannot be proved with the same 

 certainty as the orographic and climatic conditions. 



A comparison of the amounts of outflow in different years 

 shows that in individual rivers more important differences occur 

 than are generally supposed, that these differences in rivers of 

 different character and unequal force are very different in amount, 

 and that in the same river they decrease do-.vn stream. 



With regard to the difference in amount of outflow in the 

 various seasons and months, the following average values were 

 obtained. The amount of outflow in winter (from the beginning 

 of November to the end of April) is to that of summer, at the 

 parts of the stream examined, in the Rhine as I : 0-922, in the 

 Weser as 1 : 0-434, in the Elbe as I : 0-457, in the Oder as 

 I : 0-525, and further Aovm stream as i : 0-522, in the Vistula 

 as I : o 486, and in the Memel as I : 0-3S9. A better idea of 

 the regularity of the quantities of outflow is given by the relations 

 of these for the drye-t and the wettest month of the year ; in the 

 case of the Rhine this ratio is i : I -458. ia '!>« Weser i : 4, in 

 the Elbe i : 5-238, in the Oder I : 4-5, and further down 1 : 3-68 ; 

 in the Vistula i : 4-19, and in the Memel I : 4'5I. 



The causes of the difference in the ratio of the largest and least 

 monthly amounts of outflow must chiefly be sought in the 

 presence or absence of collecting basins, as also in the 

 orographic and climatic conditions. In the Rhine all those 

 fact°ors comhine which affect the regularity of outflow. It 

 possesses in the Swiss lakes large reservoirs; its river-region 

 comprises mountains of various height, and plains, so tha' the 



