ORANGES AND LEMONS IN AUSTRALASIA. 465 



equal to 0.014 per cent, of nitrogen or 1,568 pounds per acre, and that the urine of an 

 animal spent a year before might similarly increase, not only the nitrogen hut the 

 phosphates and other ash constituents of plants gathered from far and wide, it is evi- 

 dent that such a mode of sampling is liable to error and stands condemned. 



The samples of soil analyzed by Mr. W. A. Dixon were taken as fol- 

 lows : 



A piece of sheet iron -was bent into a tube, li inches in diameter, leaving a slit 

 one-quarter of au inch wide down one side, and one end of this was fastened to a 

 convenient handle. The ground was gone over at regular intervals of about a chain 

 apart, when the tube or scoop was screwed into the soil, and this being relieved by a 

 spade the section of soil could easily be removed by passing a knife through the silt. 

 The diflFerent sections allowed to dry and thoroughly mixed and sifted must give a 

 tolerably average sample of the soil of the area operated on. This may seem a tedi- 

 ous operation, but without this preliminary care the work of analysis is not only use. 

 less but often misleading — which is worse. 



To arrive at the degree of exactitude which is required in soil analyses to be of 

 value, we should consider the quantities of concentrated manures which are found 

 practically of value in agriculture and determine the quantity of the same constit- 

 uents in the soil to within such limits at least. If we take the weight of an acre of 

 soil at 1 foot deep as 5,000 tons, it will not be far from the truth as an average, and 

 this depth may be taken as that at which, being subjected to tillage, crops derive the 

 bulk of their nutriment. On this basis O.OOl^er cent, of any ingredient would rep- 

 resent 112 pounds per acre. Again, if we take the quantities of concentrated arti- 

 ficial manures which are applied as top dressing and otherwise, we find 'that a soil 

 analysis requires to be carried to the third place of decimals with certainty to be 

 comparable with practical results, for some substances at all events. For my part I 

 think that this third place is about the limit to which analyses can be carried with 

 any degree of certainty, and to do this it is necessary to carry out to the fourth and 

 take the nearest number in the third, and this requires that large quantities be 

 operated on, which require much care and patience. 



The constituents determined are those soluble in cold concentrated 

 hydrochloric acid by agitation duringforty-eight hours, and this solution 

 may be fairly assumed to contain all the constituents of the soil at all 

 readily available sis plant food. Ordinary commercial hydrochloric acid 

 was used containing already iron and sulphuric acid, and to it were 

 added small quantities of bone ash dissolved in acid sulphate of magnesia 

 and chloride of potassium. A similar quantity of this prepared acid 

 was put through all the processes used in each determination, and the 

 quantities present weighed, so that the numbers obtained for the soil 

 are the differences between it and the hydrochloric acid used. This, to 

 a great extent at all events, eliminates errors due to the solubility of pre- 

 cipitates in water and the re-agents used. 



The first soil dealt with was taken from Mr, H. Heard's property, 

 where the agricultural classes had been taken for the purpose of study- 

 ing how well-ordered orcharding can be carried on with profit. 



The soil on Mr. Heard's property is typical of a very large area of the 

 shale series. It is described by Mr. McKay as of a decided red color, 

 due to the presence of iron. It is a good, sound loam, seeming to be 

 more clayey in composition than the analysis actually shows. It is 

 the virgin soil. The geological formation here is Hawkesbury sand- 

 stone, under shale beds, which vary in quality. 



