fruits of exp. D were brought in for analysis between 3 P.M. Oct. 6 and 3 P.M. Oct. 7. The large series E furnished samples for the chemical laboratory from Aug. 19 to Oct. 16. The two short series were grown in large pots sunk in the ground, and furnished abnormal fruits of the kind here considered from Sept. 10 to Oct. 6 (exp. A) and from Aug. 15 to Oct. 4 (exp. B). Thus, three of these collections were made over a wide period, during which the plants were subjected to great variations in meteorological conditions. Furthermore, these repeated collections from the same individuals must be composed of fruits representing the plant in very different physiological stages. Sufficient proof of this assertion is to be found in the demonstration that the proportion of teratological fruits becomes smaller in successive collections (Harris and Gortner, 1914).

This table by no means includes all of the fruits produced by the plants. Others remained on the vines after the discontinuation of collections for chemical purposes because of the lateness of the season. Furthermore, large series of fruits were dissected and recorded for morphological purposes but omitted from the chemical samples because not sufficiently mature.

TABLE I

Data pertaining to distribution of the collections from which the abnormal fruits

With regard to this last point, the greatest care was taken to secure fruits at as nearly the same stage of development as possible. In the gathering of the fruits only those with dried calyces were intentionally taken. The fruits were scrutinized again with regard to this character and to maturity of seeds as they were dissected. In general, only fruits whose seeds were on the average more than half covered by the aril were used. All the fruits which showed any indication of the whitening of the wall which precedes the development of the red color, assumed in ripening, were discarded, as were also those in which the arils had assumed very much color.

Before leaving the question of materials, it may not be out of place to indicate the relative weights of the fruit wall (freed from seeds) and of the abnormal carpellary mass. The averages for the individuals samples are given in Tables 2-6.

The average of the mean weights for the three series of observations which are large enough to make it worth while to compute the probable errors are:

Thus, the mean weight of the included mass is only two thirds of that of the fruit wall. In only a single sample does the weight of the prolification exceed that of the ovary wall.

3. Experimental methods. A. COLLECTION OF MATERIAL. In any study of cell sap it is essential that the juices be obtained in a condition as nearly as possible identical with that in which they occur in the vacuoles of the protoplasm. In gathering the fruits, every reasonable precaution consistent with the magnitude of the task which had to be accomplished was taken to (a) secure fruits in a condition free from external moisture or contamination; (b) to protect them from loss of moisture by evaporation; (c) to preserve the wall of the ovary and the included mass, after dissection, in such a way as to incur as little change as possible in the conditions of the tissues; and, (d) above all, to avoid bringing about any artificial differences between the compared samples.

To attain these ends the fruits were collected with great care and dissected as soon and as rapidly as possible. The proliferous fruits were placed, as soon as found, in a container with moistened bibulous paper and closed by ground glass or a rubber sealed cover. Additional proliferous fruits were added to the container until a sample sufficiently large for the expression of sap was obtained. The abnormal mass was not removed from the fruits until the sample was prepared for chemical analysis. The very exacting task of dissecting over 100,000 fruits would have been an almost impossible undertaking but for the untiring zeal of Miss Margaret Gavin, Miss Lily Gavin, Miss Edna K. Lockwood and Mr. Charles W. Crane, for whose assistance we desire to express our sincere thanks.

B. EXTRACTION OF THE SAP. The included mass was separated from the ovary wall. After weighing, the samples were packed separately in thick walled test tubes, which were tightly closed with rubber stoppers and, as an extra precaution, capped with oiled paper fastened around the neck of the test tube by a tightly drawn rubber band. The tubes were then plunged in a mixture of finely chopped ice and salt at a temperature of -17°, or lower (care being taken to keep the mouths of the test tubes above the freezing mixture). The freezing box was placed in a larger ice box to maintain the lowest possible temperatures and allowed to remain at least 10 hours to insure the complete freezing of the tissues (See Gortner and Harris, 1914). This method of freezing the tissues is less expensive than that by liquid air (Dixon and Atkins, 1913) and, for large samples, is much more convenient and apparently quite as effective.

The contents were removed (after being carefully thawed) with great care, to prevent any contamination, folded in a small square of heavy muslin cloth (which had been boiled through three changes of distilled water and dried at 110° in the absence of dust), and the juice expressed by means of a small, heavily tinned, "beef-juice" press. The liquid was centrifuged at high speed to remove suspended solids and the physico-chemical constants determined on the clear sap.

C. DETERMINATION OF THE SPECIFIC ELECTRICAL CONDUCTIVITY. The electrical conductance, *, at 30°, was determined in the

5 Aften ten hours the temperature of the ice and salt solutions was still as low as -8° to - 10°.

usual manner, using a Freas conductivity cell having a cell constant of 0.4119, obtained by taking the specific conductance of 0.1 N KCl as 0.01412 at 30°.

POINT.

D. DETERMINATION OF THE DEPRESSION OF THE FREEZING The depression of the freezing point was carried out by the well known Beckmann method, using the modifications we have suggested (Gortner and Harris, 1914), which make for more rapid work. The freezing point of the sap from the proliferous mass was determined immediately after that of the sap from the corresponding ovary walls. All freezing points were corrected for the concentration caused by the separation of ice due to undercooling.

- 0.0125 иΔ'

where A' is the observed depression, u is the degrees of undercooling and A is the corrected depression of the freezing point.

E. DETERMINATION OF THE SPECIFIC GRAVITY. The specific gravity at 20° of the plant sap was obtained by weighing in a small pycnometer holding 5.6830 gm. of water at 20°. The maximum error in the specific gravity determination is not greater than ±0.0002.

F. DETERMINATION OF THE CONCENTRATION OF THE SOWt. solutes

was

LUTES IN THE SAP. The concentration of the sap Wt. solvent determined by evaporating to dryness, at the temperature of a water oven, exactly 10 c.c. of the sap in a small weighing bottle. The weighing bottles were then placed in vacuum desiccators over conc. sulfuric acid and the desiccators exhausted to a pressure of less than 30 mm. Hg, and allowed to stand for at least 10 hours. They were then weighed. Drying at 105°-110° gives a smaller value for solids, but we believe that this is not caused by a further

• Cf. Abderhalden: Handb. d. Biochem. Arbeitsmethoden, I, 485-498, or any good manual of physical chemistry.

This type of cell is very well suited for rapid work. The electrodes are held rigidly in place by small glass rods so that the cell may be shaken to remove traces of moisture. The cell was frequently tested with 0.1 N KCl, but during a period of nearly a year (during which time we have made some 1500 determinations of electrical conductivity), the cell constant has not changed. Cleaning the cell with chromic-sulfuric acid does not alter the cell constant.

loss of water but by the evaporation of some volatile constituent, or else by the decomposition of organic compounds. There seems to be some decomposition even when the drying is done in a water oven. Perhaps a more ideal method would be to dry in vacuo over sulfuric acid without the aid of heat, but owing to the large number of samples with which we had to deal such a procedure was not feasible.

where s is the solids in 10 c.c. and d is the specific gravity at 20°. G. DETERMINATION OF THE MEAN MOLECULAR WEIGHT OF THE SOLUTES. The average molecular weight of the dissolved substances is given by

the calculation of which is facilitated by tables published elsewhere in this Journal: 3, p. 259-263, 1914.

4. Discussion of data. The individual constants are given in the accompanying series of tables (2-21). Numbers in the left hand columns are the laboratory numbers of the samples.

A. SPECIFIC GRAVITY AND CONCENTRATION. (Data, Tables 2-6 and 17-21.) These are the simplest possible measures of the properties of the sap of the two kinds of tissues.

It would be surprising if there were not distinct differences between the means of the specific gravities of samples taken from separate cultures and at different times. The data for mean specific gravities in the appended summary are for the three larger series:

E

Prolification

1.017831 ± 0.000156
1.018125±0.000119
1.019500±0.000247