- 23 - 



tions, as A. A. Ismailsky imagines, but as will be seen below, preemi- 

 nently from surplus water not used up by the vegetation on the surface. 



Finally a third process of water circulation side ways, in a hori- 

 zontal direction undoubtedly goes on. In order to elucidate this process, 

 the following experiment was carried ont on the Odessa experimental 

 field on a piece of waste land in August when the soil for a certain 

 depth was dry. A fine jet of water (about 2 m.m. in diameter) from a 

 wooden cask was by means of a syphon turned on to a certain spot for 

 several days. Then, on a straight line drawn through the point to which 

 the jet of water had been directed eight pit holes were dug at a distance 

 of 35 cm. from each other, and samples were taken from all the pit- 

 holes at every other 5 cm. of depth. The following picture of water 

 distribution was secured thereby. 



At similar d'slances from the source of the water, in pit holes I 

 and VIII a most layer at 25 to 35 cm. depth was found, in II and VII — 

 from 20 to 45 cm., in III — from 5 to 80 cm. But in VI, which occupied 

 an analogous position to III, from 5 to 60 cm. In IV from to 125 cm., 

 and in V— from to 120 cm. (table III). 



And as there was a distance of 122,5 cm. between the source of 

 the water and the edges of pitholes I and VIII, and as the water perco- 

 lated to a depth of 125 cm. it is evident that water was imbibed in a 

 horizontal direction on all sides for a distance equal to the depth of its 

 percolation downwards. Consequently, water diffuses itself in the so'l 

 horizontally, with the same energy as it does vertically. 



In conclusion I must admit that the propos'tions which I put for- 

 ward previous (in 1094) regarding the circulation of water, somewhat 

 differ from those conclusions presented by me now, which have been 

 arrived at by subsequent evidence. 



The real developement of „the critical horizon" as I then called it, 

 goes on precisely in that depth (50 to 70 cm.) where the formation of 

 a very large number of roots takes place; and falls in May or June, 

 when the greatest number of roots are first developed. 



Clearly fhe crackmg of the humid layer at a depth of about 70 cm. 

 does not result from any cause of a mechanical character, such as 

 the raising of water by capillary action, but simply from the absorpt'on 

 of water by the roots of plants. 



