LIGHT. 51 



tration of sap resulted in endosmosis, which increased the turgor of the 

 guard-cells and in consequence caused opening. Then, after a time, 

 reconversion into starch was assumed to take place and the stomata 

 closed. In many cases the starch is manufactured in the guard- 

 cells, but in some it is not, when they have no chloroplasts. In 

 the latter type of stomata the carbohydrates are produced in the 

 chlorenchyma and transported to the guard-cells. 



The starch-content of the guard-cells was studied in several of 

 the series, in order to determine its relationship to light intensity 

 and stomatal movement. After several trials, the following method 

 was adopted as showing most clearly the relative amount of starch 

 present in the plastids: The strips in which the stomatal openings 

 had been previously measured were washed free of alcohol and left 

 in distilled water one-half hour. They were all trimmed to the 

 same size and shape, and each placed for exactly 5 minutes in 1 c. c. 

 of N/5 solution of KI and I. The strips were then washed immedi- 

 ately in a large amount of slightly alkaline water to stop the action 

 of the iodine, and mounted in glycerine jelly. The slides were 

 compared under the microscope and arranged in a series according 

 to the amount of starch estimated to be present, taking into con- 

 sideration the size and color of grains. They were then assigned 

 numbers from zero to 100, based upon differences observable in the 

 slides, and these numbers were used for the starch index. The sys- 

 tem was crude in many respects, but was the most satisfactory one 

 available, and is believed to give a general estimate of the starch 

 present in the guard-cells. 



One attempt was made in the case of the stomata of cow-beet to 

 use a colorimeter process. Fifty stomata were dissected from a strip 

 of epiderm and placed in a capillary tube 4 cm. long and 1 mm. in 

 diameter. A solution consisting of 2 grams KI and 1 gram of iodine 

 in 1,000 c. c. of water was used to stain the grains, the action going 

 on for 2 hours. For standards, a set of 11 similar tubes were made 

 up, ranging from a KI + I solution without starch, a 1/10 dilution, 

 a 2/10 dilution, etc., to 10/10 suspension of soluble starch and KI+I 

 solution. The two solution-tubes nearest in color to the guard-cell 

 suspension were compared several times and the difference in color 

 estimated in tenths. It was hoped by this means to get a truly 

 quantitative measure of starch, since the amount in the index tubes 

 was known. But the method was abandoned, because it was very 

 difficult to get the guard-cell starch into suspension, and because the 

 differences between these suspensions were mostly too slight to be 

 detected. 



The starch index determined in this manner did not show the close 

 correlation to either light changes or stomatal movement that Iljin 

 and Lloyd have found. Lloyd observed a clear and definite relation 



