to err, and realize it, is to learn. When teaching science, we often structure the work so highly that opportunities for making mistakes are eliminated. A most significant form of learning comes from the process of making a mistake, realizing it, and then attempting to correct it. Children should be allowed to make errs on their own. There is a place in the educational pattern to devise situations which lead children into making errors.

a sequence
from the Elementary Science Study (ESS) unit being developed on melting ice cubes provides several examples of teaching by having students make mistakes. The children (4th graders) were asked to determine the melting time for ice "cubes," of various shapes, placed in water.

the first problem became one of inventing ways to make unusual shaped ice cubes in the freezer at home. Make a spherical one was particularly difficult. How would you make a round ice cube? When attempting to do this, the children made numerous false starts but were able to invent some satisfactory techniques. One child froze a snowball that had been dipped in water. Another blew out the inside of an egg thru two small holes and used the egg shell to form an egg-shaped ice cube. Balloons filled with water produced ice spheres when frozen. When two teacups are placed with their open ends together, a somewhat spherical space is formed inside. Some students thought of making an ice ball this way, but had some trouble getting water to stay in the top cup which was upside down. One student resorted using the teacup method successfully, and other who tried were curious to learn how he filled the top cup. The teacher wondered too. "That ways easy," he said. "I froze the water in one cup and then turned it upside down on a second cup of water and froze it again."

the next step was to determine how long it took for the various ice shapes to melt in water. Which will melt fastest -- ice in the shape of a cube, a cylinder, a cone, a pancake or a sphere? The children made the observations at home. When the results were compared, it was obvious that there was no consistency at all. Some children found their pancake melted faster, while others reported that their cube or sphere had disappeared first. Also , the melting times for a particular shape ranged from a few minutes to over an hour. What was wrong ? The students developed several ideas when they discussed this question.

"Maybe the ice was not frozen all the way thru."
"Maybe some children stirred the others did not."

The class decided that there should be no stirring. Several of the students thought that the temperature of the water should be controlled. "Let's all use cold water." "I think warm water would be better."

when asked by the teacher how they would tell the proper temperature, it was generally agreed that FEELING would suffice. Several groups of children were then asked to prepare samples of what they considered to be warm water. When these were tested by hand, it was found that the water temperatures were quiet different. A thermometer seemed necessary, so each child was provided with an inexpensive one to use at home for his melting experiment. It was decided that everyone would use water at 70F . Students presented several different methods of obtaining water of this specific temperature.

another factor considered was the size of the ice "cube" that was melted. There were general agreement that the amount of ice used should be the same, but there was some problem of deciding how this could be accomplished. The children were familiar with the use of a tablespoon and a cup for measuring when cooking. A uniform amount of water was making ice could be measured in this way.

the importance of the volume of water used to melt the ice was not apparent to the children. An experiment was performed in class to help the children see that some of the variations in their results had undoubtedly been caused by the different amounts of water in which the ice had melted. The problem was to compare the time for the ice to melt in water with its melting time in air. A previous experiment had shown, much to the surprise of all, that ice cubes in air take about an hour and a half to melt. The paper cups which were provided for melting the ice in water ranged in size from 4 ounces to 32 ounces. This would insure that the children would unsuspectingly use varying amounts of water. When it was noticed that the ice cubes in the largest cups melted faster, it seemed that the speed of melting depended upon the size of the cup. The children could guess roughly how much ice was left in various sized cups without looking. A little additional thought indicated that the melting rate was related to the amounts of water rather than the size of the cup.

now, armed with considerable knowledge of the factors which affect melting, the children repeated the original experiment of melting ice of various shapes at home. They attempted to use uniform piece of ice, melted in identical volumes, of water, and at the same temperature. Even so, their results were not consistent, but they were better.

one should not be disappointed by the children's failure to obtain meaningful results. What they find out about the techniques for conducting an investigation is of great value. In he lessons described above, the children had practice in designing equipment, devising techniques for experimentation, measuring, keeping records of their results, and analyzing the data which was collected. It is also important that they had many chances to make mistakes. By so doing, children learn much about the nature of scientific inquiry.