140 PLANT PHYSIOLOGY. 



by side. To enclose areas of the leaf so that only a single color will 

 reach them, rectangular openings were cut in a piece of velvet to match 

 the smooth faces of the bottles; this was then glued to them in such a 

 way that the joints between the bottles were covered by the velvet and 

 the leaf was exposed only under the flat face of each bottle. The other 

 side of the leaf was covered by black velvet glued to a strip of glass. 

 Under this arrangement the iodine test showed that in a day nearly as 

 much starch was made under the red screen as under the white one 

 (distilled water), while a very much less, but still appreciable, quantity 

 was made under the blue, and none at all under the green. To prevent 

 the corks being forced from the bottles by expansion of the liquids 

 under heat, a thread between cork and neck connected the air-space over 

 the liquids with the air outside. 



8 (page 94). For strict accuracy, allowance must be made in this 

 apparatus, not only for temperature, but also for vapor-tension and 

 barometric pressure, concerning which corrections see note 2 above. 



9 (page 98). There are other methods of subjecting seeds to growth 

 conditions without oxygen, such as surrounding them by a neutral gas 

 (hydrogen), but the following gives perfect results: Take two of the 

 slenderest test-tubes, and, heating them near the neck in the flame, 

 draw them out until the neck is but 2 or 3 mm. in diameter; when cool 

 slip into each two or three soaked oats from which the glumes have been 

 removed. Into one of them put a bored rubber stopper holding a glass 

 tube, to which attach a tube from an air-pump, such as the Chapman 

 nump, and exhaust thoroughly; while the exhaust is still on heat the 

 narrow neck with a bunsen flame until it seals itself entirely. Then treat 

 the other similarly except that, before sealing, the air is to be admitted 

 (the exhaust is used on the latter also in order to be sure that the ulti- 

 mate difference may not be due to something connected with use of 

 the exhaust). The two tubes are then put aside in a favorable growth- 

 temperature. Instead of the test-tubes ordinary glass tubing sealed at 

 one end and drawn to a narrow neck near the other may be used. 



Another very ingenious method, which I have repeated with entire 

 success, is that given by Murbach in School Science, I. 25. It is, how- 

 ever, much more effective to use a control experiment than simply to 

 test the vitality of the seeds later, as Murbach recommends. The con- 

 trol would be treated in every way like the principal tube except that it 

 should be allowed to cool, and hence readmit air, before being sealed. 

 The perfection of the vacuum in such tubes may subsequently be tested 

 by holding the sealed end under \vater and cracking it off with nippers, 

 when the height to which the water rises in the tube will give a measure 

 of the vacuum, which can be made nearly perfect by this method. Oats 

 give excellent results. One must be careful not to explode the tube in 

 sealing it, which may be guarded against by using the same flame to boil 



