608 



SCIENCE 



[N. S. Vol. LIV. No. 1407. 



iigured by Eiley, and the membranous disk 

 is shown correctly with the sheaths inside, 

 corresponding to the beginning of the bor- 

 ing process. But the position of the abdo- 

 men is impossible; indeed at this stage, 

 when the disk is formed, the abdomen is 

 held not only vertically but even bent for- 

 ward to some extent above the thoras; and 

 at no time during the whole process is the 

 ovipositor inserted as far behind the insect 

 as drawn by Lutz. Like Comstock, Lutz 

 shows the wings in a vertical position and the 

 antennse are held obliquely upward which is 

 possible but not characteristic. Mention 

 should be made that Eiley too, already gave 

 a picture, undoubtedly from a preserved speci- 

 men, of 'the extended membrane, the two 

 sheaths just leaving it, as would be the case 

 as soon as the membrane begins to collapse. 

 This illustration shows very well how the 

 ovipositor at the beginning of the process is 

 held in a vertical direction by being sunk 

 into a ventral furrow of the abdomen, which 

 renders its basal portion quite invisible. 



It becomes a matter of interest that, of 

 many authors commenting on such a familiar 

 insect as our large, long-tailed ichneumon 

 ily, and on its oviposition, only comparatively 

 few have watched the process long enough to 

 verify its details, and that, in fact, some of 

 these details have never been clearly estab- 

 lished though Megarhyssa is common in 

 many localities. Does not this indicate that 

 we have been neglecting the ecological for the 

 systematic aspect of entomology? 



Werner Marchand 

 Mendham, N. J. 



A CONDENSATION PUMP 



Condensation pumps of the following par- 

 ticular type have been used in our work for 

 a number of years and the design seems to pos- 

 sess sufficient advantages over others in both 

 simplicity and compactness to merit this note. 



The method of operation of this pump, in 

 which the exhausting process is accomplished 

 in two stages, will be made clear by reference 

 to the cut. In the initial or " rough " stage, 

 A, the mercury vapor is ejected at relatively 



high pressure from a small nozzle into a long 

 narrow throat. The nozzle opening is made suf- 

 ficiently small that the pressure of the vapor 



Pig. 1 



in the boiler, instead of being practically 

 limited to 2 or 3 millimeters, as in the case of 

 the ordinary vapor pump, may attain a value 

 of 75 millimeters or more depending upon the 

 heating. The efficacy of this arrangement 

 was first pointed out by Stimpson.^ The evac- 



1 Washington Acad. Sci. J., 7, pp. 477^82, Sept. 

 19, 1917. 



