"When you live in my part of California," writes Caggiano, "interest in the growth of plants comes easily. Nearly everything will grow here. Having hit on the idea of doing an experiment involving growth-promoting substances, I found that I could easily get lots of help and advice, particularly from Kenneth C. Jones of the Department of Botany and Plant Biochemistry at the University of California at Los Angeles.

"Since cantaloupe was available as a source of young seeds, I selected it for the experimental subject. Seeds were separated from enough melons to make about 200 grams, fresh weight. The seeds were chopped and then soaked for 24 hours in a 500-milliliter solution consisting of four parts of acetone and one part of water at a temperature of five degrees centigrade. The solution was poured from the seeds and the procedure was repeated twice. The three solutions were combined and the acetone was evaporated under vacuum at a temperature of 30 degrees C., leaving some 300 milliliters of water and the extracted materials from the seeds.

"This solution was then partitioned, a process involving the separation of acid, neutral and basic components. The pH, or acidity, was first measured by conventional techniques. The solution turned out to be somewhat acid. The pH was increased to (slightly basic) by the addition of a few drops of weak potassium hydroxide solution. At this pH gibberellins are insoluble in ethyl acetate but soluble in water.

"The solution was then placed in a separatory funnel. Approximately 100 milliliters of ethyl acetate were added and the funnel was shaken for two minutes to mix the ingredients thoroughly. The water, being immiscible and denser than the ethyl acetate, settled to the bottom. The stopcock of the funnel was opened and the water was collected in a flask. The collected water was then similarly processed four more times. After each partitioning the ethyl acetate was discarded. The pH of the water solution was next decreased to 3 by the addition of sufficient hydrochloric acid. In this highly acid state the gibberellins are soluble in ethyl acetate. The acid solution was then partitioned with ethyl acetate. The water fraction was reextracted five times and the ethyl acetate fractions were combined to give a total of 500 milliliters of ethyl acetate solution. The solution was transferred to a beaker and evaporated overnight in a fume hood. A filmy residue of plant extract remained in the beaker.

"I then used a technique of elution chromatography to purify any possible gibberellins, or gibberellin-like substances, contained in the residue. The apparatus consists of a cylindrical glass funnel that has an inside diameter of 30 millimeters and is 45 millimeters high. The funnel plugs into a companion filtering flask below. The side arm of the filtering flask connects to a vacuum pump. The outlet of the funnel is closed with a plug material such as glass wool.

"I soaked four grams of a mixture of equal volumes of powdered carbon (Norit A, decolorizing carbon, neutral, prepared by the Fisher Scientific Company of Fair Lawn, N.J.) and Celite, an inert diatomaceous silica (prepared by Johns-Manville), for three and a half hours in a solution of one part of hydrochloric acid to 25 parts of distilled water (by volume).

"The carbon-Celite slurry was then poured into the chrolrratographic column, and the mixture was washed three times with distilled water, which was removed by suction. After the third washing the water that had collected in the vacuum flask was discarded. Next I dissolved the plant extract in a 50-milliliter solution that was half acetone and half distilled water. The acetone was thereafter evaporated from the solution and enough water was added to increase the volume to 50 milliliters.

"That solution was added to the top of the chromatographic column and was passed through the carbon-Celite mixture by turning on the vacuum. Gibberellin-like substances adsorbed by the carbon-Celite mixture remain in the column but many of the other ingredients pass in solution to the flask below. This solution was discarded.

"The next procedure releases the growth-promoters from the column in controlled amounts. Twenty-one separate solutions of acetone and distilled water are made up in 10-milliliter units. The first unit consists of five parts of acetone by volume in 95 parts of water; the second, 10 parts of acetone in 90 parts of water; the third, 15 parts of acetone in 85 parts of water and so on in increasing increments of 5 percent acetone per unit to the 20th, which consists of 100 percent acetone, as does the 21st. Each unit of solution, beginning with the 5 percent acetone, is washed through the column consecutively and collected separately. The 21 separate fractions are then labeled and placed under a fume hood for evaporation to dryness. Each of the dry residues is then dissolved in one-milliliter solutions of 50 percent acetone and water, by volume, for application to the test plants.

"As test plants I used two genetically different dwarf mutants of corn (Zea mays). The first mutant, known as dwarf-5, responds to all nine gibberellins. The second, dwarf-1, responds to a different degree to some of the gibberellins and was used during a subsequent growth experiment that will be explained. I grew the seedlings at room temperature in the presence of light. "I applied each fraction to a set of 10 plants, which were 10 days old, using .1 milliliter on each plant. The extract was applied by medicine dropper to the surface of the first unfolding leaf of each seedling. A group of 10 untreated plants was grown under identical conditions but was reserved as a control. Ten days following treatment the amount of growth of each plant was recorded by measuring to the nearest millimeter the length of the first two leaf sheaths.

"The measurements were then statistically reduced and plotted as graphs. To make the reduction I first found the total length to which both the first and the second sheath had grown for each plant. The sums of the lengths for each group of 10 plants were then added together, and the average sum of the lengths of the first two sheaths was determined for each group. In the case of the control plants, for example, the combined growth of both the first and the second leaf sheaths of all 10 plants was 558 millimeters, for an average of 55.8 millimeters. The deviation in the growth of each plant from the average growth of the group was computed by adding together the differences of the sums of the growth of the first and second leaves from the average growth of the first and second leaves. For instance, in the case of the control group the first and second leaves of the first plant grew to a length of 65 millimeters, 9.2 millimeters more than the average obtained from the measurements of 10 plants.

"In the case of the controls the sum of the squares of the individual deviations amounts to 921.6. This figure is used to compute the standard deviation, which is equal to the square root of the quotient that is obtained by dividing the sum of the squares of the deviation of each plant, in this case 921.6, by the total number of plants minus one, or in the case of the control group N/921.6/9 = 10.1 millimeters. Finally, the standard error of the arithmetic mean is computed by dividing the standard deviation (10.1 in the case of the control group) by the square root of the total number of plants in the group: 10.1/N/10 = 3.2. Computations similarly made for each group of plants were then tabulated and the data used to make a graph of the growth response. These statistics were used to indicate whether differences in the two groups were meaningful or just due to chance.

"Six separate experiments were made. The accompanying table [Figure 1], listing the data of the first experiment, is representative. The second experiment was made to verify the results of the first. The third and fourth experiments were similar to the first two but used mutant dwarf-1 corn to help identify, at least to some extent, the presence or absence of specific kinds of gibberellins. The fresh weights of young cantaloupe seed extracted for the experiments were, beginning with the first experiment, 196, 201, 211 and 206 grams respectively. The accompanying graphs [Figure 2] give the results.

"Experiments using identical procedures, except for the extraction, were also made on both mutant dwarf-5 and dwarf-1 with pure gibberellic acid for comparing the response to that of cantaloupe extract. The results are shown in the accompanying graphs [ Figure 2].

"The experiments suggest that the cantaloupe contains 'gibberellin-like' substances. The plant extract produced a growth response in the two genetically different dwarf mutants that appeared to be identical with that of the known gibberellins. The peaks of the graphs, in addition to indicating the high activity of the extracts, suggest that at least two, and possibly more, gibberellins are present in cantaloupe. The fact that the peaks occupy approximately the same positions on the graphs of dwarf-5 and dwarf-1 also suggests that the same growth-promoter is active on both mutants.

"Finally I estimated the total amount of gibberellin-like activity. The seedling response to each fraction of the extract was compared with the response due to pure gibberellic acid. Then the amount of gibberellic acid that would have been necessary to cause the response to the extract was determined. The total GA3 equivalents were multiplied by 10 because each value represented an average of 10 seedlings. The resulting estimate indicates that the extract applied to dwarf-5 mutants amounted to the equivalent of 40.8 micrograms of gibberellic acid per 397 grams of cantaloupe seed, or .1 microgram per gram. On dwarf-1 mutants the estimate was that the extract applied was equivalent to 45.2 micrograms of gibberellic acid per 417 grams of seed, or .11 microgram per gram. In other words, 86 micrograms of gibberellic acid would be required to produce the activity observed during the four experiments."

Some of the most entertaining experiments have been invented by compulsive doodlers, individuals who cannot sit with folded hands and minds in low gear during idle moments. Roger Hayward, who illustrates this department, is such a person; he has generated many off-hour inventions.

A recent letter is illustrative. "Some years ago," wrote Hayward, "I bought a set of steel balls graded in size from one inch to 1/2 inch in steps of 1/16 inch. I wanted them for use as gauges. As the occasion arose I added four more one-inch balls to the set, and over the years a number of smaller ones just seemed to accumulate. I also happened to have an old oval metal tray about 20 inches long. During a quiet evening a few months ago I picked up the tray and for no particular reason started to roll some of the balls inside the rim. Suddenly it occurred to me that here was a chance to learn something about how balls bounce off each other. I recalled from my student days at the Massachusetts Institute of Technology that it is difficult to calculate the change in velocity when two bodies collide, except in certain simple cases.

"It also came to mind that if one ball falls a certain distance, measured vertically, it acquires a certain velocity, no matter whether the fall is vertical or down any sloping path. Barring friction losses, the ball will have the same velocity at the bottom of the path regardless of the slope. I calibrated my tray by drawing a series of horizontal lines from rim to rim one inch apart, with a vertical center line. (A china-marking pencil is good for the job because the 'lead' is made of soap and the marks wash off easily.) I taped a small bubble-level to the center of the tray so that the center line could be kept in the vertical plane [see Figure 4]. I needed an indicator for marking the excursion of the balls. A good one was made of a cork sanded fiat on one side so that it would not roll. A rolling ball will push the cork up the slope and the cork will stay there. The cork absorbs some energy from the rolling ball, of course, but the resulting error can be minimized by rolling the ball several times. Eventually the cork is pushed to its end point, where its position can be recorded by the china-marking pencil. I found that a single ball of one-inch diameter would eventually push the cork to about six-sevenths of the distance the ball had rolled down the other edge of the tray, indicating a friction loss of about a seventh of the total fall.

"The next experiment was made to find out how much of the energy of a rolling ball is transferred to a similar ball with which it collides, as shown by the accompanying illustrations. When the missile ball was rolled from a height of six inches, the target ball rose to only 1 1/2 inches, as measured on the slope. Obviously not all the energy was transferred. The experiment was then repeated with a pair of one-inch balls centered at the bottom. The No. 2 ball now rose somewhat higher than the single ball. When three balls were tried, the one at the outer end of the string, No. 3, failed to rise as high as had No. 2 in the previous experiment. With four balls, No. 4 rose higher than No. 3 but not so high as No. 2. Part of the energy of the missile ball appeared to reside in its spin; the spin energy was not fully transferred. In fact, when I used only one target ball, the spin that was transferred may well have inhibited the transfer of other energy and thus have accounted for the twin-ball performance.

"To test this supposition four balls were placed in contact at the bottom of the tray, each marked with a meridian arc, all arranged in parallel. Th end ball was prevented from moving by finger pressure, and the missile ball was allowed to strike the target ball of the string several times. Six drops of the missile ball were enough to prove that the impact caused the balls to rotate in alternate directions by amounts that diminished toward the restrained ball. This observation made sense of the previous results. Rolling energy is a significant part of the total energy of the rolling ball, and it is transferred at lower efficiency than the energy of translation.

"Next I started putting sets of different-sized balls in the tray A small ball absorbed only part of the energy, the missile ball continuing along after it. At this point a somewhat more sophisticated apparatus appeared to be desirable, so I improvised one from a drawing board, three blocks of wood, a thin metal strip and a set of clamps. With straight slopes for sides, rulers can be used to measure the rise and fall of balls. The steel strip that serves as the track must fit rather closely to the wooden back strips, particularly where the path curves, or inordinate amounts of energy are dissipated.

"With this new apparatus I repeated the first experiment, except that I used target balls smaller than one inch. A target ball 15/16 inch in diameter rose a little higher than the one-inch target ball, and a 7/8-inch ball went even higher. As I added smaller and smaller balls to the array the gains increased. With seven balls in the string the ball at the end, 9/16 inch in diameter, rose higher than the starting point of the missile ball. A 1/2-inch end ball climbed three inches higher than the missile ball's starting point. When I added a 5/16-inch ball to the end of the string, I had to lower the starting point of the missile ball to keep the end ball from jumping off the track and lower it still more for an end ball of 3/16 inch. With the starting point of the missile ball only 3 1/4 inches above the base line a 3/16-inch ball climbed 11 inches!

"Now I had to revise my thinking. There was no likely loss of total energy that could not be accounted for, but in the light of the low efficiency of energy transfer from ball to ball I had assumed that the system was substantially like that of a falling fluid. A small stream of water flowing from the low point of a large reservoir will rise no higher than the level of the reservoir. I had failed to see in the case of the balls that the energy is transferred by an elastic wave; the amount of energy transferred is not necessarily limited by the velocity of the impact. It would seem that a row of balls of graded sizes acts like an electric transformer, if one thinks of the mass as the current and the velocity as the voltage. By electrical standards the efficiency of the ball-transformer leaves much to be desired, of course.

"There is also a hydraulic analogue in the old, time-honored and largely forgotten hydraulic ram. In this device a large pipe filled with flowing water is suddenly blocked by a self-closing valve. The impact of the fluid mass drives part of the water into a separate chamber, compressing a volume of trapped air. When the flow stops, a valve closes and the trapped water is forced by the air to a substantial height. I spent the final moments of the evening experimenting with a simplified version of the hydraulic ram: a short length of rubber tubing that acted as a siphon between an elevated pan of water and the kitchen sink. At the discharge end of the tube I inserted a small piece of metal tubing with a .042-inch hole in the top. When I stopped the flow of water with my thumb, a narrow jet squirted more than six feet high out of the hole. The springiness of the compressed air in the ram is provided by the rubber; the springiness of the steel balls, by the elasticity of the steel."

Bibliography

EVIDENCE FOR GIBBERELLIN-LIKE SUBSTANCES FROM FLOWERING PLANTS. Bernard O. Phinney, Charles A. West, Mary Ritzel and Peter M. Neely in Proceedings of the National Academy of Sciences, Vol. 43, No. 5, pages 398-404; May, 1957.

THE LIFE OF THE GREEN PLANT. Arthur W. Galston. Prentice-Hall, Inc., 1964.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.