TEST VARIATIONS.

Shannon News, 7 February 1928, Page 3

TEST VARIATIONS.

WHY DO FACTORY AND ASSOCIATION TESTS VARY?

SOME OF THE CAUSES

(Contributed.)

Ever since the pay out of dairy factories has been based on the fat content of milk or cream supplied, it seems to have been the supplier’s privilege to grouse at his test--it is always too low, never too high. The days of systematic Group Herd Testing arc here, and apart from other benefits derived from the testing of the individual members of a herd, the average test and total butterfat of the whole verifies in a very definite way the tests obtained at the factory, always providing that the Association’s sam oles were obtained from the herd under normal conditions.

The two tests, that is. the factory milk test and the herd’s average test as obtained by the herd tester, cannot be expected to compare to a point, and the reasons for this are puzzling some members of the Herd Testing Associations. Where the difference is great doubt is cast on the accuracy of the work of either testing officer or manager. The tester’s average, except in .rare cases, will fce higher than the factory test. The variation ranges from nothing to .4 or even .5, but for the average herd of 50 cows .2 is the usual variation. The same applies to compav'son between total fat on the Association’s records as compared to fat actually supplied. In comparing records it is this total fat ' and not the average test that this, after allowances have been made should be taken into account. The Association calculates the fat produced by the whole herd over 30 days and where ncessary, • should be in actual comparison with the fat as calculated by the factory to have been supplied over a similar period. This is by' far the more accurate basis, of comparison but even this will be found to vary’ at Limes and with some herds regularly. NOW FOR THE REASONS.

First and foremost comes the problem of water in the milk delivered to the factory. Under the writer’s observation this will be up to three-parts of a kerosene tin to a 20-gallon can. In many cases a certain amount of water for flushing out machines is allowed to reach' the cans, and if this is but a very small proportion of the whole no exception can be taken to such a practice, but with care it should not be tnoie than a pound or two and the resulting difference in test will be negligible, though with- other reasons may mount up the whole. It is not here intended to stress the inadvisability of .catering milk, but the man (one cannot calf him a dairyman) who does so in these days of co-operation, is robbing both himself and his neighbour, .n wet. districts, where milk is left Hoove red overnight, rain also plays an important part in test variation, the samples taken ana tested by the held tester are pure milk —-they contain no ulded water. The following may prove of interest o those who do water milk. It is the esult of many actual experimental ests: — Unwatered. Watered 51b. Watered 101 b.

Prom this it will be seen that very ittlo water per can is necessary to ■nake a difference of a point whcie the ester’s average is 4. The separator will have the same ■ffcct on the comparison if it, is not skimming clean, .though this, of course, s carelessness or ignorance rather. than vilfulness. Herd testers will willingly est- out any member’s separator, and ,y ’ finding "a robber separator, many have paid’their testing fees with,part .f the money saved from cream fed to i,lgb ' ORDINARY CAUSES.

While water can play such an imjortant part in test variations, we ait ;noi e concerned with the ordinary causes ' leading ' to variation. Prom practical experience as a testing offi':cr, the writer knows that in many beds a great deal of extra care is takem m getting the last drop on test milkings. It is, of course, only natural, Hint the dairyman wishes his cows to do their best, so ne sits stripping for i Title longer than usual. Srtirpings are by far the highest testing part of he cow’s How and by obtaining only a little more of this' rich milk per cow he test for the whole of the milk ob- , ained on such occasions must be rela.ively higher than is usual, and as the rat is calculated from these tests and •eckoiied for the wliole mouth by multiplying by thirty, a decided difference a fat production can be obtained. Jhe same applies to any unusual factor on test days, feed, weather or other outside interferences. If fair and accurate individual records are desired, it is ABSOLUTELY ESSENTIAL

that the herd and individuals be given regular and noil-preferential treatment on test-milkings and immediately before The weather then remains as tin uncontrollable factor. Another item, and one tluT helps the faetorv over-run to some extent, is that thmk,’tough laver of cream found on the top of the night’s milk in the morning. A careful man may never see this on his milk, for he keeps it m a void and fresh location and stirs it last thing before he goes to bed, and perhaps on a hot night sprinkles a veiy fine sprav of water over each can and thus prevents the thick layer from forming. Thick and tough cream is never found on a rainy night where the cans are outside. The water keeps the cream from actual contact with the evaporating air. In the morning this skin is churned into small pieces through-out the milk by means of a stirrer, but the pieces as compared with the original state of the fat globules are very large. It.is a minimum reckoning to say that half of this cream skin is lost to the farmer and goes into the factory over-run. It is imposiblo

for one of these small lumps to reach the test bottle. Some skim this cream and put it in with the warm morning s milk, which is certainly a much better method than merely stirring it m, but even so a great deal is lost. Sexual disturbances at times will be the cause of much variation. This factor cannot be overcome effectively m large herds but what can I>c done for the individual records is done. Where a large number of cows are so affected the total returns and average of the herd must bo adversely affected. I-leavilv frothed milk is the bugbear of anyone wishing to obtain an accurate sample. If a sample of pure froth is tested' it will compare very unfavourably with the test of the milk below, so that a sample rvitli more than its fair proportion of froth will give a test unfavourable to the cow. This may be the reason why the average of small hand-milked herds is usually so much closer to factory returns. OTHER REASONS.

The question of cream or milk taken for household purposes, milk for the calf, of strippers, and others not tested and so not in the Association averages and totals; of warm milk poured into cold in the cans; of fresh samples tested by the testing officer as against stale, preservatisod samples by r the manager. These, and other causes perhaps of no great, importance in themselves, but each helping a little to tilt the balance, must be taken into consideration when comparing individual tests month, to month of Association records with factorv. COWS ARE NOT MACHINES.

Probably no farm animal approaches the highly specialised dairy cow in the delicacy' of some of her vital physiological processes. She is reactive to an extraordinary' degree. She is largely a creature of nerves, which control the organs in milk making. The play of heat, cold, light, darkness, rain, drought, sight, sound, and odour upon her nerves, and the influence of maternal instincts upon her behaviour and functions, must inevitably' be expressed in variations of her butterfat output from day- to day-, regardless of the kinds or amounts of food she, consumes. The lesson that observation should instil is that the dairyman who knows his business will manage liis cows not as machines but as extraordinary', sensitive, changeful and feminine individuals with some of the inscrutable idiosyncraeies t,hat distinguish females of a wellknown higher species. The Dairy Division lias issued a bulletin on test variation the concluding paragraph of which is—“ The foregoing information is put forward with a view to assuring dairymen and breeders that when variations occur in tests of butterfat, it should not be immediately assumed that the variation is due to carelessness, lack of ability, or malice on the part of the person conducting the test.” It remains for each dairyman who has his cows under test, or intends so to have, to consider all these 1 ieas oils, and perhaps others known to himself, and he will find that the troubli> of varying tests lies not in inaccurate testing of either tester or manager, but more likely with conditions over which neither have control.

Test. .< , per 1001b. per 1001b. 3.0 ' 2.86 2.73 4.0 3.8 3.63 5.0 ■ 4.76 4.54