Figure 6, at the end of the sixth day, shows still more plainly the germ undergoing a gradual change in the egg, enlarging and assuming a darker hue. The outline of the circle is now gradually acquiring the form of an ellipse, and in a live embryo the line of demarkation should be distinct. If it is at all wavy and irregular in its outline, and instead, remaining intact, the contents of this ellipse show a disposition to assimilate with the surrounding liquids when the egg is revolved, it can be safely removed as a dead duck. [Illustration: FIGURE 6.--EGG AT END OF 144 HOURS.] Figure 7 represents a dead embryo, as it will appear from the seventh to the twelfth day. The germ being separated and appearing in dark irregular blotches over the entire surface of the egg; the egg having become nearly opaque over its entire surface. At this stage the egg, if it has not already become so, will soon be very offensive. These should be removed at once and handled carefully the while, as they are apt to explode and unpleasant consequences ensue. The operator should run no risks, as discoloration on the outside shell of a duck egg is a sure sign of decay, and they can safely be taken from the machine. There are always a certain number of duck eggs (especially during the month of August and the latter part of July) that have the appearance of fertility during the first three or four days of the hatch, but do not possess vitality enough to carry them through. These die at all stages of the hatch; neither operator nor machine is responsible for them. This is caused by the condition of the mother bird. [Illustration: FIGURE 7.--A DEAD EMBRYO.] In order to economize the room thus made by the removal of the fertile eggs, I run a small 150-egg machine, in connection with twenty-one of the largest size, using it, as it were, as a tender. When filling one of the larger machines, I always fill one tray in the smaller one so that when the eggs in the large one are tested, after the third day, there will usually be eggs enough in the small tray to replace those removed as infertile, so that the large machines are kept full during the entire hatch by the little one. Thus the small machine is made to accomplish far more than it would were it run through the hatch. I am thus enabled to have a hatch come off nearly every day, consequently our eggs are never older than that when introduced into the machine. Always date each day's quota of eggs--keep them by themselves, then there will be no mistakes made. I have known parties to keep one general receptacle for their eggs, and when filling their machine take them from the top, while the bottom ones were never disturbed, not even turned, and of course soon became worthless for any purpose.