December 2, 1921] 



SCIENCE 



553 



The rod is then screwed to the end of the 

 tube of the injection apparatus by means of 

 the screw point in which is a fiber washer to 

 make the joint tight. The rod is then clamped 

 in a mechanical pipette holder, either that of 

 Barber or one described in an article al- 

 ready printed. The next step is to fill the 

 pipette with mercury. To do this open the 

 stopcock and see that the rubber tubing con- 

 nected with the stopcock is full of mercury. 

 With a strong clamp close the tubing about 

 four inches from the stopcock. Along this 

 four inches place several screw clamps which, 

 on being screwed down, will produce sufficient 

 pressure to drive mercury almost to the tip 

 of the pipette. The stopcock is then to be 

 securely shut off. 



We are now ready for action. Squeezing 

 the metal tubes by one or other of the lever- 

 age clamps will drive mercury through a 

 pipette having an aperture of only one 

 micron (.001 mm.) in diameter. Move the 

 pipette by means of the pipette holder till 

 its tip projects into a hanging drop of the 

 solution to be injected. Release pressure on 

 the steel tube and some of the solution will 

 be drawn into the pipette. ISTow lower the 

 pipette and move the moist chamber till the 

 cell to be injected is brought into view. The 

 pipette is now raised until it punctures the 

 cell. On applying pressure to the steel tube 

 the solution is readily injected. The appara- 

 tus may also be used to withdraw materials 

 from the cell. 



The apparatus is extraordinarily sensitive. 

 The meniscus of the mercury in the pipette 

 responds instantly to the pressure of the 

 leverage clamps. A comparative estimation 

 of the quantity of injection material used may 

 be made by focusing, first, on the mercury 

 meniscus, then on the tip of the pipette and 

 measuring the distance of the two focal points 

 by means of the fine adjustment screw of the 

 microscope. 



A more complete description of this appara- 

 tus will shortly be published. 



EoBERT Chambers 



Cornell. Medical College 



ON THE EMISSION AND ABSORPTION OF 



OXYGEN AND AIR IN THE EXTREME 



ULTRA-VIOLET 



Up to this time very little has been known 

 of the spectrum of oxygen in the region of 

 wave-lengths shorter than A2000. Some pre- 

 vious investigators were unable to . obtain a 

 spectrum in this region. " No lines or bands," 

 says Lyman, " were observed between A2000 

 and 1230." ^ Schumann, however, had suc- 

 ceeded in photographing some continuous 

 maxima of which the most refrangible has 

 a wave-length of about 1850 Angstroms. 

 Moreover, Lyman had observed that the great 

 absorption band of oxygen diminishes in in- 

 tensity as it approaches A1230, but he thinks 

 that another absorption band exists " lying in 

 the region shut out by the absorption of 

 fluorite." This preliminary investigation 

 was undertaken, therefore, to test the emis- 

 sion and absorption of oxygen and air in the 

 region of wave-lengths shorter than those 

 transmitted by fiuorite. 



The apparatus used consisted of a vacuum 

 grating spectrograph, containing a Rowland 

 concave grating of 50 centimeters focus, about 

 15,000 lines per inch, ,'>nd a ruled surface of 

 approximately 2 inches. A discharge tube 

 of internal capillary, end-on type and with 

 aluminum electrodes was employed. The tube 

 was also provided with a quartz window for 

 Hg comparison spectrum and opened through 

 a slit directly into the receiver. A method 

 has been developed of making Schumann 

 films, and these were used for the spectro- 

 grams. Commercial oxygen, dried with phos- 

 phorus pentoxide, filled the receiver and con- 

 nected discharge tube to a pressure of about 

 0.4 mm. When the spectrum of air was ob- 

 tained, this gas was likewise dried and filled 

 the receiver to about the above pressure. The 

 time of exposure varied from 20 minutes to 

 2 hours for the gas spectra, while an ex- 

 posure of 3 minutes was found to be suffi- 

 cient for the Hg-arc comparison spectrum. 

 The apparatus was so arranged that both the 

 first and second orders of the Lyman region 



1 Lyman, ' ' The Spectroscopy of the Extreme 

 Ultra-violet, ' ' p. 82. 



