Agricultural Chemistry — Slieej) Feeding and Manure. 279 



they soon became surfeited, a fact which is not to be wondered at, 

 considering that the substance contained about 12 per cent, of 

 nitrogen, a circumstance which we shall see as we proceed would 

 be much in favour of its use, so far as the resulting manure is 

 concerned. 



Sheep were next taken, which, from their docility and manage- 

 able size, are convenient subjects for experiment; and, owing to 

 their very general utility, they seem better fitted than any other 

 description of farm-stock to be assumed as their type, in an in- 

 vestigation of the general chemistry of the production of meat and 

 manure, though at the same time it must be admitted, that it 

 would in some respects have been desirable to have included 

 bullocks also in the inquiry. It being essential to our object to 

 collect, without loss and free from litter or extraneous matter of 

 any kind, the whole of the excrements of the animals, both liquid 

 and solid, to be accurately weighed and sampled for analysis, pens 

 were constructed in a spacious barn, each about 8 feet by 1\ feet, 

 and having a flooring of rafters, on the plan proposed by the Rev. 

 A. Huxtable, the width of the rafters being about 3 inches, and 

 the distance between them about f inch, so as to allow the whole 

 of the excrements to pass through. Below this flooring, which 

 is raised perhaps 2 feet or more above the ground-level, sheet- 

 zinc is fixed at such an incline as to allow the urine to run off, 

 while the dung is retained upon it. The urine passes through a 

 spout into a covered pail kept constantly underneath to receive it, 

 and to which a straining basket is fixed to stop any solid portions 

 that may come down, the dung being removed from the zinc at 

 pleasure for weighing, sampling, &c. The food, of course, was 

 also accurately weighed. This arrangement appears sufficiently 

 simple, and well calculated to attain the end desired, but we shall 

 see as we proceed, that in the details of practice difficulties are 

 met with from which serious errors in result may easily arise, 

 unless great care be taken to avoid them. These sources of error 

 are chiefly connected with the great difficulty of obtaining samples 

 for analysis which shall accurately represent the bulk of matters so 

 heterogeneous and variable in their composition as the food and 

 excrements of animals ; and when it is remembered that in some 

 processes of analysis a few grains only of substance are operated 

 upon, and that from the composition of these that of the whole is 

 calculated, it will readily be understood, that in an investigation 

 like that in question — which itself constitutes indeed from the 

 beginning to the end one process of quantitative analysis — the 

 greatest care is requisite, if erroneous conclusions are to be 

 avoided. And, that the reader may be enabled to decide as to the 

 legitimacy of such as we shall found upon the results, a full de- 

 scription of the experiments will be given, which moreover may 



