successfully a series of such lights, beginning with a two-wick oil lamp behind a large Florence flask filled with water, the flask serving at once as a condenser and heat filter. Afterward, in succession, was used an argand gas lamp; a Welsbach burner; a three-wick projection lantern, burning camphorated oil; a 50-candle power incandescent lamp; and a four-burner acetylene lamp with stereopticonFig. 2. Apparatus Complete for making Instantaneous Pictures of Living Microscopic Animals. double condenser, to the final and most efficient light of all, an arc light which can be regulated between 1,000 and 5,000 candle power. Of course the arc light gives out a great amount of heat as well as a satisfactory quantity of light, and were the rays to fall directly upon the microscope lenses the temperature would rise high enough to endanger their mountings. To obviate this difficulty a distilled water cell two inches thick is placed in front of the stereopticon double condenser, which is directly before the arc at such a distance as to parallelize the divergent beam. It is a common idea that an alum solution is the best for absorbing heat while transmitting light. That such is not the case is proved by experiment. Melloni and others have shown that distilled water will intercept more heat rays than a solution of alum. The writer has verified that conclusion and is also able to show that distilled water transmits more light, has practically the same heat-absorbing power through a considerable range of temperature and also has the advantage of no formation of crystals in the cell. Distilled water is necessary, as with common water the air bubbles collect on the parallel faces of the cell as the temperature rises, and obstruct the passage of light.

The form of apparatus shown complete in Fig. 1, using a Zeiss microscope and Bausch & Lomb adjustable camera with automatic regulating arc light, gives excellent results for ordinary dry mounted