price of feed in the different herds in Newaygo County Association shows. a difference in the price of linseed meal for the same month of $5.00 per ton. These herds were only a few miles apart and it is evident that if the feed had been bought in a co-operative way at least $5.00 per ton could have been saved to the owners of herds e, D, G, Q and R.

The price fixed for pasture varies considerably in the different herds, according to the kind of pasture and the time of the year. As a rule, however, when we take into consideration the value of the land and the crops which it would produce the price is undoubtedly much too low except in the case of wood-lot pastures. Thus in some cases the cost of summer feeding is calculated as only one-quarter that of winter feeding, a relation which is out of keeping with the experience of our most expert dairymen, especially on land which costs from $60.00 to $100.00 per acre.

The cow tester has been able to be of much assistance to the individual dairyman in the way of formulating rations. Thus in the case of herd f of the North Ottawa Association a comparison between the first and the last month of the year shows a decided improvement in the profits in spite of the fact that the price of feed on the whole had increased and the price of butterfat had decreased 42 cents.

TABLE SHOWING INCREASE MADE IN THE PROFITS OF ONE HERD PER MONTH AS A RESULT OF MORE JUDICIOUS SELECTION OF ECONOMICAL FEEDING STUFFS.

It will be seen that although the price of feed had gone up the cost of the ration of the 11 cows had been reduced almost 50 per cent and the profits increased from $2.21 to $16.87.

In this connection should be mentioned the advisability of holding a meeting of the members at least once a month during the winter months, when questions of feeding can be discussed. Such discussions will prove of great benefit to the members in an educational way and will in time develop a co-operative feeling among the members and this indeed is a valuable asset for any community.

BREEDING AND SELECTION.

If the individual dairyman wishes to improve the dairy qualities in his herd or provide for a herd which will give him a better income year after year he can go about it in two ways. He can by the aid of his records from the cow testing associations, select his best cows to head the herd and by breeding them to good sires with dairy qualities be reasonably certain to obtain young animals with dairy qualities. The heifers of such breeding are used to replace the poorer cows in his herd and in a few years he will have a herd which will produce much more net profit for the same labor and feed. The second way is to buy from reputable breeders dairy cows of known qualities or young heifers of

good breeding. But the price asked for such animals, while it may not be exorbitant, is usually more than the average farmer is willing to pay and he will find that if he wants good dairy cows he must raise them himself. And his success will depend on his realization of the great law in breeding that "like begets like or the likeness of an ancestor." Just as surely as the good dairy cow, as a rule, produces offspring with good dairy qualities, so surely must he expect offspring without dairy qualities from the poor dairy cow. For this reason he must choose only his best dairy animals for mothers of his future herd. The systematic dairyman fixes a minimum production as a standard of excellence in a mature cow, and while one is content if a cow produces him 250 pounds of butterfat in a year, another wants 300 pounds and selects only cows capable of such production for his mother cows. In every case the cow must be bodily sound.

But the certainty of inherited dairy qualities in the offspring does not depend upon the good qualities of the mother alone. We can only be reasonably certain when she has been bred to a sire whose ancestry shows the same good qualities which we wish to produce. When the dairyman buys the bull he has at the same time chosen the breed he wishes to work with as well as the type he wants to reproduce in his herd, and for this reason he should choose with care and deliberation. In case the herd is made up of cows of mixed breeding the bull will exert an influence on the offspring of more than 50 per cent. But as with the cow so with the bull. His value is not known until he has been tested. One must see his offspring and know their qualities before one can know his value, and many bulls have been killed before their value was known. Members of a cow testing association have an opportunity to exchange bulls whenever it is necessary to dispose of them, and in this way the ruthless slaughter of young bulls in the prime of their usefulness is prevented. When the type has been established in a herd it is also desirable that the sire and the dam be as nearly alike as possible, in which case the sureness of their ability to reproduce their good qualities in their offspring is increased and retrogression is guarded against. All of which goes to show that it is necessary to have a fixed plan if one shall realize the most out of his work for better animals. The improvement is greatly retarded if the cows one year are bred to a Jersey bull and the next year to a Holstein or Shorthorn or any other breed, and improvement cannot be expected at all if the cows are bred to bulls of unknown breeding or ancestry. That sort of breeding or crossing should be discontinued and the sooner the better. Breeding within a certain breed is always to be preferred.

If the members of the associations will follow these rules they will be certain of permanent improvement. An improvement as was made in the herd of Aug. Kinck (Bulletin 137) is a worthy goal for any dairyman and should be within reach of all. Mr. Kinck increased the average milk production per cow in 7 years from 7,300 to 11,333 pounds per year and the average production of butterfat per cow from 245 to 401 pounds per year, and this improvement was all accomplished by putting into practice the fundamental principles of breeding above outlined in connection with his records from the cow testing association.

Many of the members of the cow testing associations in Michigan have

responded very promptly to the lessons cow testing teaches. Thus during 1907, the first year's existence of the North Ottawa Association, 41 cows were discarded principally because even a short test or perhaps . an increased interest in their business in connection with a short test revealed their unprofitableness.

Below follows a record of the cows disposed of during the year and since the close of the year's work many more were sold.

While it is recommended that a cow be tested at least a year before her value for dairying is determined, it has been possible in many instances to tell even after a short test that it would be unprofitable to keep her. However, it is urged that all factors be taken into consideration before a cow is condemned as cows do not do equally well every year.

During the year, in addition to a number of purebred heifers, 33 cows were bought by the members of the Newaygo County Association and of these 21 were purebred dairy animals or high grades, and while only one man owned a purebred dairy bull when the association started, 22 such bulls are now found in the herds, a fact which promises well for the future development of the herds. It was not expected that the second year's average would show much increase over the first year's average as the association has not been organized long enough for the results of the better breeding to be shown in the yield, and the cows purchased were in nearly every instance heifers, which calved in the fall. However, the average butterfat production per cow of the herds tested the first year increased 8 pounds per cow, in spite of the high feed prices and a natural tendency on that account to save on the feed. The average cost of feed the first year was $29.28 per cow and for the second year $36.42. The profit per cow the first year was $20.99 and for the second year $27.43 per cow. This increase in the profit was made possible by the increase in the price of butter.

The profits and returns for $1.00 expended are figured on the basis of the yield of butterfat and the price of butterfat is the price received by the farmer at the creamery. This price, with the same market varies 3.3 cents within the same association, and a comparison between herd O and herd i of the Newaygo County Association tells the reason. It

877370

is simply the difference in the price of butter where summer dairying is practiced as compared with the price received under winter dairying, and is a strong argument in favor of winter dairying.

If the cow comes fresh in the fall she gives the most milk when the price of butter is high and when the dairyman has the most time to feed, milk and care for her, and when she is let out on the pasture in the spring she flushes up on the yield of milk almost like a fresh cow. She goes dry during the hottest months in the year and when her owner has the least time to take care of her.

If she comes fresh in the spring she gives a wonderful flow when the pasture is luxuriant but when the pasture gets dry she drops off a little; when the first frost comes she drops off a little more and when permanent cold weather comes she usually goes dry. In the case of herd O and with the cows to freshen in May, the daily yield from 8 cows during December was 43 pounds or an average of 5 pounds a day for each cow.

The cost of roughage is based on the local markets and for the concentrated feed is the actual price paid for it.

In determining the total profit, it has been assumed that the value of the calf, the skim milk and the manure from each cow would pay for the labor in feeding and caring for her. This is believed to be liberal pay, and perhaps too liberal where the calf is a high grade or from purebred dairy parents.

COW-TESTER'S NOTES.

Leading dairymen have for a number of years advocated the weighing of individual cow's milk at regular intervals in order to get a reasonably accurate estimate of her yearly yield. If this is not done and the farmer is content with weighing the milk off and on as he happens to think of it, the impressions he gains are apt to be misleading. Some cows give a wonderful flow of milk when they are fresh and if the milk is weighed then, the estimate of her yearly yield is apt to be too high. Then again, that kind of a cow may soon fall off on her milk and perhaps go dry three-fourths of the year. On the other hand, a cow which gives a moderate flow of milk and keeps it up the year round may give the largest yield in a year. If the milk is not weighed regularly a farmer will remember especially the large flows and the cow which gives the moderate flow will not get her just due in his estimation.

In the cow-testing association the weighing of the milk usually begins in the afternoon. The cow-tester weighs the evening's milk and the morning's milk and adds the two figures together to get the amount given for a day. Likewise a proportionate sample of milk is taken at both milkings for testing. It should be strictly observed that the milking is done at the usual milking hours so that he weighs neither more nor less than one day's milk. If the milking hours are so arranged that they extend over more than twenty-four hours, the weight of the milk will be too high and vice versa; therefore, the milking should be done at the regular milking time and the cows should be milked in the same

rotation as the previous day. The cows should also have been milked clean at the previous milking. These factors have much influence on the weight and the test of the milk on the testing day.

When there is competition between the herds it is customary for the tester to come the day before so that he can guarantee that the cows have their usual allowance of feed, are milked clean and are milked at the regular hour. If it is impossible for him to come the day before he can compare the amount of milk on testing day with the amount sold or used on the previous days. If one of the competitors then milks his cows comparatively early the day before testing day it will usually result in a larger yield on testing day but a lower test as it has been found that the greater the distance between milkings the more tendency to poorer milk. And if on comparison the amount of milk on testing day is found to be much larger than the yield on previous days the tester is justified in reducing each cow's milk record proportionately until the same agrees with the average for the previous days. In that event the test would be too low and the farmer would have accomplished the opposite of what he intended. In the cow-testing associations the farmers are cautioned to observe regular milking hours and if they lack the understanding of the importance thereof, it becomes the duty of the cow-tester to inform them. It is a good plan for the cow-tester to vary the testing day slightly or alter his route slightly occasionally so that he sometimes comes unawares.

Before milking the cows' numbers or names should be entered in the stable book in the same rotation in which they are to be found in the year book. As it is one of the purposes of the work to find the actual status of the man's dairy business, every cow in the herd should be entered whether she gives milk or not. Of course heifers are not entered until they have their first calf. The milk is weighed in what is called a shot gun can and should weigh even pounds in order to avoid mistakes.

This can should be eight inches in diameter and twenty-inches high and hold thirty-five pounds. It should also have a rim around the bottom or a handle near the bottom so it may be easily emptied. This can should be part of the cow-tester's outfit. In this way mistakes in subtraction of the weights of different pails are avoided and time is gained in the stable, as the milker can pour the milk into the shot gun can and go to milk the next cow without waiting for the cow-tester to weigh and sample.