40 



KEPOKT OF CHEMICAL LAHOKATOKY 



Centrifugal 

 method 



Size of the 

 clay particles 



be set accurately to the minute will enable him to devote a little time to other work during 

 the whirling. We have found an electric alarm the most reliable for the purpose. One of 

 the forms of centrifuges used in America is shown in Fig. 4. The machines are well known 

 and need no special description. It is desired simply to call attention to the method 

 of controlling the results. It is usually stated that this must be carried out by the use of 

 the microscope. While this is, in a measure, true, it would be useless to expect to obtain 

 results exactly comparable with those from sedimentation, in such a manner. The 

 slightest variation in the size of the particles from the standard which has been adopted 

 would, in many soils, make a marked difference in the results. It must be recollected that 

 the particles are never spherical, so that exact measurement by means of a micrometer 

 is practically impossible, and no two observers could be expected to get precisely the same 

 result. The plan we have adopted is to treat several soils of different character by 

 the sedimentation method and then to determine the length of time of whirling in the 

 centrifuge, run at a known speed, to obtain the same result. 



If the soil has been sufficiently puddled at the start we have found that the number of 

 whirlings required is much lessened, almost the entire amount of clay being removed after 

 the sixth decantation. In the machine at hand, 6 minutes has been fixed upon as the 

 period of whirling, the' number of revolutions being about 840 per minute. The machine 

 figured runs at a much higher speed and the length of whirling would be materially 

 reduced. 



It has been found that after the 6th or 7th whirling, when nearly all the clay has been 

 removed, the remaining soil packs so loosely at the bottom of the tube that the swirling of 

 the liquid, due to the arresting of the machine, causes some of the fine silt to be drawn 

 up into it. As the result of this the water never becomes quite clear, and the end of 

 the clay removal is difficult to judge. This may be avoided by increasing the time 

 of whirling, after the 6th or 7th, to 10 minutes, so that the soil residue packs more closely. 

 The error introduced by the longer whirling at this stage of the process is negligible. The 

 difficulty mentioned would probably not be experienced with machines adapted to higher 

 speed. 



The size of the daij particles. Different workers have adopted various limits for the 

 size of the particles taken to represent the group designated " clay." The most 

 common upper limits are '002 mm. and '005 mm. Hall' employs the first mentioned and 

 separates the clay by 24 hours' subsidence from a column of water 8-| centimetres in height. 

 Kilroe, Seymour and Hallisy- employ a method which is attributed to Hall and in which 

 the height taken is 15 centimetres, the particles being stated to have a maximum diameter 

 of -005 mm. Hilgard^ describes a method of subsidence of 24 hours' duration, the colunni 

 of water being 20 centimetres in height "whereby all grain sizes of and above O'Ol mm. 

 diameter are removed from the turbid liquid." Atterberg-* recommends either the limit 

 ■003 mm. obtained by a 4 hours' subsidence from a column of 10 centimetres, or of '002 mm. 

 from a column of the same height but with the time of subsidence increased to 8 hours. 



It is obvious that the above statements as to the size of the particles are highly 

 discordant. We have made many measurements, all the results of which point to the 

 fact that the figures given by Hilgard and by Kilroe, Seymour and Hallisy and many others, 

 are not even approximately correct. Our measurements agree more nearly with those 



' The Soil, 2nd Ed., p. 51. 



-' The Geological Features and Soils of the Aiiriealliu-nl Slatioii of the Dq)t. of A'jricullare at 

 r.allnhaiiu-, 1910. 

 ■■■ Soih, p. 89. 

 • ijher die KlassiJikatioK der Budenkorner, A. Atterberg, Kaliiuir. 



