PRATT: CYCLOLOMA ATRIPLICIFOLIUM. 107 



had been dissolved out. All the globules which stain like oil dis- 

 solved out in sections lett in xylene, chloroform or ether for forty- 

 eight hours. Certain of these globules stain evenly, but most of 

 them stain unevenly. Usually a spherical core (ranging from 4 to 

 20 per cent of the volume of the globule), located in the center or 

 at the side of each unevenly stained globule, stains a darker or 

 lighter red than the rest of the globule. Two separate spherical 

 cores were present in each of certain globules observed. 



Saponification tests on fresh sections did not indicate the pres- 

 ence of saponifiable oil in the globules. The Tunmann reagent 

 used for the tests was made up fresh, and was tested on castor- 

 bean sections, the oil of which saponified. The test was repeated 

 on freshly cut sections of stem, and more reagent was made avail- 

 able to the mounted sections by placing small strips of cork at the 

 edge of the cover-slips, which were then sealed with wax (65 per 

 cent beeswax and 35 per cent rosin), or by the use of hollow- 

 gi'ound slides. The slides were placed in the electric oven and 

 examined daily. The globules, which stain like oil, dissolved out 

 in from twenty-four to forty-eight hours, and no saponification 

 was evident in sections examined with the aid of the microscope 

 or polarizer. 



The oil globules in sections steamed for eight hours and then 

 left in Sudan III or alcannin for twenty-four hours stained like 

 the globules in sections, newly cut from formalin material, left in 

 either of the two stains for twenty-four hours. Evidently no 

 substance in the globules escapes when the sections are steamed, 

 for the globules average the same size as the globules in sections 

 not steamed. Condensed vapor from steam, which had been 

 passed over sections and then condensed on cold petri dishes, did 

 not stain yellow when exposed to the vapors from heated iodine 

 crystals, and the presence of viol atile oil was not demonstrated. 



The above tests indicate that an oil is present in the globules, 

 but evidently not a volatile nor a saponifying oil. 



In many of the cells containing globules staining like oil there 

 are also greenish-yellow globules, which do not dissolve out after 

 the sections have remained seven days in xylene, chloroform or 

 ether, and which do not stain when the sections are left in Sudan 

 III or alcannin for forty-eight hours (fig. 9, r'). These globules 

 were tested for glucosides as follows: Sections were boiled in 

 5 cc. of Fehling 's solution, and at first there was no change in the 

 color of the reagent, nor could crystals of cuprous oxide be seen 

 in these yellowish-green globules with the microscope; a precipi- 



