He said:  “The sand tray early in the season gave the best hatches and most vigorous chicks we had, but later on things got too wet and the chickens drowned.”  No nicer demonstration of science in practice could be desired.

In the present-day incubator of either type we are wholly at the mercy of sudden climatic changes of vapor pressure.  For the slower changes from season to season some control by greater and less amounts of supplied moisture, or by ventilator slides is available, but little understood and seldom practiced.

It will certainly be of interest to my readers to know the actual hatches obtained with the prevailing type of box incubator.  By actual hatches we mean the per cent. of live chicks taken out of the machine to the per cent. of eggs put in.  The ordinary published hatches, based on one per cent. of fertile hatches, are a delusion and a snare.  When eggs are tested out many dead germs come out with them and the separation of microscopic dead germs from the infertile egg is, of course, impossible.  Such padded and show hatching records do not interest us.

Where incubators are run on top of the ground I have found the results to be poor and to improve, the bigger and deeper and damper and warmer and less ventilated the cellar is made.  The reason for this is plain.  In such a cellar the vapor pressure of the air is not only greater but is less influenced by the shifting vapor pressure of the outside air.  In a good cellar the operator, though his knowledge of the factors with which he deals is grievously deficient, learns, through long and costly experience, about what addition of moisture or about what rate of ventilation will give him the best results.  In the room more subject to outside influences, the conditions are so constantly changing that uniformity of practice never gives uniform results, and hence the operator is without guidance, either intelligent or blind, and the results are wholly a product of chance.

As proof of my contention I may give results of a series of full season hatches for 1908, each involving several thousand eggs.

First, a state experiment station, the name of which I do not care to publish.  Incubators kept in a cement basement which has flues in which fires were built to secure “ample ventilation.”  This caused a strong draft of cold, dry air, making the worst possible condition for incubation.  The hatch for the season averaged 25 per cent. and was explained by lack of vitality in the stock.

Second, Ontario Agricultural College.  A room above ground, moisture used in most machines and various other efforts being made to improve the hatches by a staff of half a dozen scientists.  Results:  Hatch 48 per cent.—­incubator manufacturers call the experimenters names and say they are ignorant and prejudiced.

Third, Cornell University:  dry ventilated basement representing typical conditions of common incubator practice of the country.  Results:  Hatch 52 per cent., results when given out commonly based on fertile eggs and every one generally pleased.