out coating to preserve its integrity, is cut off at the desired depth and removed for weighing. Many cautionary measures have been given to secure a high order of exactness for the determination. These methods are open to two serious objections. They are suited only to fairly homogeneous soils of fine texture. If the soil is composed in considerable part of gravel, or if it contains larger fragments of stone, cutting tubes or prisms can not be driven into the soil without disturbing its normal texture, and blocks of correct form and plane. sides can not be secured for weighing. Second, these methods are so laborious that they can not well be used for repeated determinations made on the same soil to check against errors arising from variations in its uniformity of texture. For use in a soil through which flinty fragments, of one or more inches or greater diameter, are irregularly but rather frequently scattered, the writers have therefore adopted a method quite different in principle from those of the classes above-mentioned. By this method, the soil is broken and removed from a roughly measured space and weighed before and after air-drying. The volume of the space from which the soil has been removed is measured by careful determination of the volume of dry sand required exactly to fill it. From the data for soil weight and excavation volume, with any necessary correction for change in the volume of the sand resulting from its transfer from the graduate to the excavation, the apparent specific gravity can be computed. The following details have been employed in the use of the method: (1) The soil was examined at a time when it was fairly dry, but sufficiently moist not to crumble too readily. (2) The surface was cleared of stubble and smoothed by use of a sharp trowel or knife. (3) By means of a rule and a knife, a rectangle was marked off upon the smoothed surface. For the soils studied, the dimensions 9 x 4 inches were chosen, so that sufficient space for the use of excavating tools might be secured. For coarser soils, larger excavations would be preferable. (4) Two straight-edged pieces of wood were laid parallel, close to and on opposite sides of the lines marked on the surface, and a vertical cut 2 or 3 inches deep was made by use of a sharp knife or flat trowel. A helper kept the pieces of wood firmly in place during the cutting, and the trowel was withdrawn very carefully to avoid any displacement of the surface soil. (5) The soil within the excavation was then removed by aid of a narrow trowel and transferred to a receiver. A piece of oil cloth was spread between the edge of the excavation and the receiver so as to catch and preserve any soil particles that might spill. The excavation was continued to approximately the desired depth by the use of a knife or trowel, in such manner as to leave undisturbed such of the larger stone fragments as were firmly fixed in the walls of the excavation, but so as to remove with the soil other fragments that came away loosely from the sides or bottom. (6) The soil thus removed was promptly air-dried. (7) The volume of the excavation thus made, was determined by filling it with sand from a graduated cylinder. The sand was delivered uniformly by pouring from a height of 2 or 3 inches above the surface and along the major axis of the rectangle. From time to time as the filling proceeded, the central ridge of sand, formed in the manner described, was leveled and filled into the corners of the excavation and into the hollows in its sides that were caused by the breaking of stone fragments. This leveling and distribution was accomplished by means of a straight-edge, with which the sand was stroked as gently as possible to avoid unequal compression. Care was taken to deliver the last portions of the sand a very little at a time, so that no excess might in any case be used for the filling. (8) To reduce as far as possible the error in the measurement of the sand, which was free from all but traces of loam, the liter cylinder was filled each time in precisely the same manner. The sand was delivered into a small funnel, set in the neck of the cylinder. When filled almost to the mark, the sand was leveled by gently rocking the cylinder without jarring it. The last portions of sand required were then allowed to trickle in from the hand. (9) To secure the highest practicable exactness of measurement, the sand remaining in the large cylinder after the excavation was filled, was transferred in like manner to a 50 cc. cylinder to determine its volume. (10) Finally, to determine what correction, if any, was necessary for a difference in the space occupied by the sand in the measuring vessel and in the excavation, a standard cylindrical brass half-peck measure (4409 cc.) was repeatedly filled with sand in the same manner in which the soil excavations were filled, and from the liter cylinder employed in the field measurements. The excavations were carried down to subsoil, which was found at depths varying from 3 to 7 inches. The uncorrected volumes of the excavations ranged from 1658 to 4552 cc., with an average of 3467 cc. The depth of the measure was therefore approximately that of the average excavation. The quantities of sand severally required for 10 fillings of the half-peck occupied on the average 4459 ± 0.77 cc. in the glass graduates; that is, the sand was less compact in the half-peck in the proportion 4409: 44590.77 or as 1000 1013. The measures obtained as the volumes of the excavations were therefore divided by the factor 1.013 to correct for the relatively greater compactness of the sand in the measuring cylinder. In the use of this method, it is necessary to determine the correcting factor corresponding to the filling material used and the conditions of filling maintained in each series of studies. The labor thus required is not great. On the other hand, the excavation can be made and its volume determined in little more than an hour, so that duplications of the determination at various points of the surface in question can easily be made. The importance of such duplication will be discussed in another paper. NITROGENOUS COMPOUNDS IN SOILS1. J. K. PLUMMER (State Department of Agriculture, Raleigh, N. C.), Associate Referee. The work outlined this year has been a continuation of that done by C. B. Lipman2, the plan of which follows: PLAN OF WORK. The problem was to test out on the same material the official Kjeldahl method for nitrogen determination in soils; the official method, as given under "Fertilizers"4; and the Hibbard method for nitrogen determination in fertilizers, as modified by C. B. Lipman for soils. HIBBARD MODIFICATION OF THE GUNNING METHOD. Place the soil in a 500 cc. or, better, in a 800 cc. long-necked Kjeldahl digestion flask, add 30 cc. of sulphuric acid and approximately 10 grams of a mixture prepared by grinding together and thoroughly mixing 10 parts of potassium sulphate; 1 part of ferrous sulphate; part of copper sulphate. Immediately shake the mixture of acid, salt and soil so that no soil remains untreated by the acid. Then digest it, first with a low flame, and then with a strong flame for 1-2 hours, depending on the amount of organic matter present. After digestion, dilute the mixture, transfer it to 1 liter copper distillation flasks, and distil into N/10 hydrochloric acid. Titrate the acid in the usual way, using either methyl orange or cochineal as an indicator. METHODS TO BE TESTED. (1) Official method under "Soils". (2) Official method under "Fertilizers". (3) Hibbard modification of the Gunning Method. SPECIAL INSTRUCTIONS. Try all of these methods on each soil as follows: Use methyl orange or Ten gram portions of soil (20 cc. of sulphuric acid). Twenty gram portions of soil (30 cc. of sulphuric acid). The period of digestion should be 24 hours in every case. cochineal as indicator. Take special care in neutralizing the acid before distillation. Preferably employ Greenbank's lye and make up by dissolving 1 part in 2 of water. Use N/20 hydrochloric acid amd N/20 ammonium hydroxid, or N/20 sulphuric acid and N/20 sodium hydroxid. Report the results in terms of cc. of acid, in mg. of nitrogen, and in per cent of nitrogen in air-dried or water-free soil. Samples of Durham sandy loam (1 A) and Iredell loam (2 A) were washed free of nitrates. To portions of the original Durham sandy loam 0.02 (1 B) and 0.04 (1 C) per cent nitrogen, as sodium nitrate, was added. Portions of the original Iredell loam were similarly treated and the results appear in the table as 2 B and 2 C. All samples were oven-dried before being sent out. The following table gives the results obtained: 1A 1A per per per per per per cent per per cent cent cent 0.017 0.015 0.025 0.028 0.032 0.028 0.021 0.025 0.019 per 1B 20 0.026 0.026 0.025 0.029 0.026 0.033 0.038 ...0.032 1C 10 1C 20 0.047 0.059 0.046 0.050 0.059 0.046 0.053 0.059 0.015 0.022 0.025 0.063 0.042 0.020 0.026 0.023 0.053 0.061 0.061 0.057 0.058 0.054 0.058 0.063 2C 20 0.052 0.062 0.056 0.057 0.037 0.036 0.037 0.039 0.037 0.039 0.032 0.043 0.037 0.033 0.046 0.039 2B 10 0.034 0.037 0.039 0.039 0.0370.038 0.039 0.045 0.027 0.041 0.043 0.039 0.036 0.039 0.032 0.038 0.031 0.041 0.054 0.048 0.052 0.043 0.068 0.045 After a careful examination of the results obtained by the different analysts, it does not appear wise to offer any recommendation for the adoption of a new method to supplant the present official method. However, the results clearly show that there is little choice between the methods now in vogue for measuring small amounts of nitrogen in soils. The official method to include nitrates does not recover the nitrogen which has been added in the form of sodium nitrate. Sometimes this method gives higher results than the other modifications tested, and sometimes not so high. Taking the results as a whole, the Hibbard modification gives about as high figures as either of the other two methods. Considering the ease of manipulation of digestion and distillation, the Hibbard method seems to be preferable. Nitrates should be determined on another sample by either the colorimetric or reduction method. No consistent difference is apparent whether ten or twenty grams of soil are taken for analysis. REPORT ON THE LIME REQUIREMENT OF SOILS. By W. H. MACINTIRE (Agricultural Experiment Station, Knoxville, Tenn.), Associate Referee. The work outlined upon the problem of lime requirement has been along two lines: (1), an effort to ascertain the conceptions and viewpoints held by those who have given particular attention to the problem with a view to defining the term "Lime Requirement"; (2), a study of one or more representative types of the several procedures advanced. It is well known to the members of this association that there exists a marked diversity of opinion as to the nature of the phenomenon causing the decomposition of calcium carbonate applied to soil. By some it is held that true, if peculiar, acids occur in soils, the hydrogen ion concentration of which may be determined. Others hold that the decomposition of carbonate is effected through physical absorption of the calcium ion, while a third conception is that the original basic silicates have undergone hydrolysis with subsequent leaching of the hydrolyzed products, thus leaving a complex mixture of what may be considered as true acid silicate salts. It has been hoped first, to reach an agreement as to the terminology; second, to arrive at some definite conclusion as to what may be demanded of a method and what procedure most nearly fulfills the exactions decided upon and, furthermore, as to whether the procedure should be considered solely as a laboratory measurement of a physical or chemical phenomenon or whether the chemical data should be susceptible of interpretation into field practice. This, of course, involves the question of the possibility of correlating lime absorption measurements with plant response. This again raises the questions of pot studies v. field studies; selection of plant indicators; purity; hardness; porosity and solubility; and proper time for, and frequency of, application. Most of these considerations involve both the chemical and biochemical factors direct and indirect, as well as the economics of the problem. To quote aptly, B. L. Hartwell, in correspondence with the associate referee, writes as follows: It seems to me that what we need is a definite criterion of what we are attempting to accomplish, by which we may judge of the merit of the rather confusing number of methods. The acquiring of more data without some definite standard is, I fear, unlikely to mean an advance. It seems to me that we must ask the question -lime requirements for what? For what kind of a crop? And requirements for how long a time, etc.? Queries as to the advisability of the utilization of pots have developed the fact that by many it is held that this method may be considered |