The chemical analysis showed a presence of 14 per cent. of phosphate of lime. The liquids remaining in the bowls, as well as the water used in washing, should be carried away at once in order to prevent the development of microbes.

When the radiator stopped we took the gross weight of the butter grains and butter-milk, afterwards stirring up the whole with a long-handled spoon for two minutes before working it up dry.

The butter-milk was weighed and the butter made in rolls of 1 to 2 kilos, to be placed on oak-rib shelves in a case with lid, which could be filled with ice in the summer. Finally, twenty-four hours later, the butter was again worked up and weighed.

D.

The following table gives the results of four experiments-A, B, C, and

On the figures of this table the following observations can be made :The time required to start the radiator, when the driving power is ready, would be about a quarter of an hour, and any excess of this would be attributed to inexperience at the start of our experiments.

Butter came out of the exit pipe one minute after the milk was sent into the radiator, but in reality only 15 to 18 seconds are required for separating and churning when the radiator is in full speed. Though we could not strictly maintain the temperature constant at 70 degrees centigrade, the small variations in the result do not seem to have been affected by this circumstance.

The cooling of the milk in passing through the separator varied between 7 and 10 degrees centigrade.

The not very important variations in the quantity of milk passed through per hour depend principally on the time lost in starting and regulating, which would be the same for a large or small quantity. You may estimate the quantity practically at 650 litres per hour (= 130 gallons).

The gross weight of butter-mixture produced which may be considered equal to the cream (when the churner is not used) is about 10 per cent. As regards the percentage of butter obtained there seemed to be a great uniformity.

The variations in richness of fat from the different specimens of milk used were very insignificant, as proved by the following analyses, as well as by the subsequent ones executed according to the excellent methods of our eminent director, M. Dudaux :

After passing through the Radiator the analysis of the separated milk shows:

In examining from these tables the only really important figure, which is the percentage of fat obtained, we establish the fact that the radiator extracted 95 per cent. of the total fat existing in the fresh milk, basing it on an average of 3.77 per cent. total fat in the fresh milk, and 0.2 per cent. remaining in the separated milk.

Analysis of the butter made showed the following composition:

In consequence of the large percentage of fat in the buttermilk, it ought to be passed again through the Radiator at the next operation.

The results which we have stated, with some short considerations, authorise us to form the following conclusions:

From a mechanical point of view the Radiator is a very ingenious and perfectly constructed machine, which works with irreproachable regularity. Notwithstanding its having no special foundation, and running at a great speed, it works without noise or vibration. The erection is very simple, at any rate to the hands in our creameries. The students at the school learnt the mechanism and work in two trials.

The starting is easy, the preparations and precautions being identical with those required for any centrifugal machine. A motor of 4 to 5-horse power is needed for the radiator which we used, and though we only had a 2-horse power boiler, we advise one of 3-horse power to heat the Pasteurisator.

From a trade point of view, the radiator offers indisputably great advantages.

The space necessary for installation is only 25 metres square (equal to about 83 feet square), including room for all necessary manipulations. Hand-labour is reduced to a minimum. Two men are quite sufficient to look after it, one for the engine and boiler, the other for the radiator and other appliances. Afterwards they both share the work of cleaning and working-up. It can be worked either as separator only and give Pasteurised cream, which can afterwards be acidified by adding lactic ferments, or at any moment during the work it can be changed to separator and churn. The result, when working it as a separator, is as satisfactory as is practically possible, and we know of no superior centrifugal in that respect.

The churning is as complete as possible, and removes in its working the anomalies incidental to acidification of cream.

At none of our comparative experiments has the actual production of butter been surpassed by any process used.

The whole of the milk is Pasteurised, separated, and the cream churned and sheltered completely from contact with the air, which is a great desideratum, and which no other method of manufacture permits.

You have also the power to treat milk only heated to 30 degrees centigrade; it is sufficient for this purpose to merely moderate the admission of steam in the Pasteurisator. During the working, the Pasteurised cream is instantly reduced to a suitable temperature, which you maintain constant by regulating the admission of cold water. This does away with the necessary previous

heating or cooling of the cream, and the defects in the butter caused by elevation of the temperature, which are invariably produced when you use ordinary churns.

Not only is the work continuous, but it is possible to manage all the details and vary them rapidly during the work within the following limits :— 30 to 70 degrees centigrade for separator and 12 to 20 for churn. The quality of the butter leaves nothing to be desired, notwithstanding that the cream has not been soured, besides it is independent of the professional skill of the workers. The butter is homogeneous, fine-grained, has good consistency, a most delicate flavour and aroma, and thanks to the Pasteurisation will keep longer than any butter made by the ordinary process, as well as being a safe food, which is all the more precious as this article cannot be sterilised. In conclusion we can say that the radiator furnishes an economical and splendid work with a minimum of buildings, material, manual labour, and time. It will consequently be adopted in all factories of any importance, particularly in those which work for the export trade.

Illustration Fig. 3-showing arrangement of buttermaking machinery, page 820-A. Tube for separated milk; B. Inlet tap, for unskimmed milk; C. Tap for chilled water; D. Refrigerator for separated milk; E. Radiator; F. Pump for separated milk; G. Return pump for chilled water; H. Tube for the unskimmed milk; I. Barrel for the unskimmed milk; K. Pasteurisation apparatus; L. Cold water cistern; M. Tap for emptying Pasteurisator; N. Framing and table; O. Pipe for chilled water; P. Cistern for the fresh milk; Q. Accelerator of speed; R. Pump for fresh milk.

A