Since I have shown that the metabolism is proportional to the sur face and Mayer has shown that the loss in weight is proportional to the volume (weight), the composition of the cassiopea must change during starvation. In other words, it loses weight faster than it burns protein (or other organic matter), and hence the concentration of the protein must increase. Mayer (1914) found the cellular layer did not decrease in thickness during starvation, and Hatai found the percentage of nitrogen to the total body-weight increases during starvation and is also greater in small than in large, well-nourished cassiopeas. There fore in attempting to calculate the metabolism from the loss in body substance, we should make it proportional to the loss in surface rather than loss in volume, since the loss in living matter seems to be propor tional to the loss in surface and the cassiopea seems to have no other important store of food than its own protoplasm, the mesoglea appar ently functioning chiefly as a skeleton.
Since the surface is proportional to the two-thirds power of the vol ume, we may assume that the protein is proportional to the two-thirds power of the weight (the density remaining practically constant). The protein equals 5.16 per cent of the two-thirds power of the weight (calculated from Hatai's data on the assumption that protein is 16 per cent N). The weight at the beginning of starvation was 100 grams and the protein 1.107 grams; at the end of one day the weight was 94.4 and the protein 1.07, being a loss of 37 mg. of protein. If we assume that 1 mg. of protein is equivalent to 4.4 cal., the metabolism the first day would be 163 calories, although I found it to be 336 to 342 calories.
Although these calculations are only approximate, since starvation is a little greater the first day than calculated by the formula, this great difference indicates that the burning of protein does not account for all of the heat. Since living cells contain lipoids or lipo-proteins and carbohydrates or glyco-proteins, it seems probable that proteins, carbohydrates, and lipoids are burned. The mesoglcea has not been analyzed separately, but is largely sea-water, with possibly a trace of glyco-protein. It probably has little calorific value, since the use of a store of food would cause a relatively greater metabolism in large starv ing cassiopeas than was actually observed.
Since the lining of the alimentary tract is not at the surface of the cassiopea and 02 must diffuse through at least a millimeter of tissue to get to it, it was decided to pull off the manubrium and study the metab olism of the umbrella. The umbrella is disk-shaped, is covered on both
sides by epithelium, and pulsates, thus circulating the water. The wound made by removal of the manubrium is of small area and is covered by an epithelium within a few hours, and the umbrella will live as long as a starving cassiopea. Some rough determinations indicate that the respiration of the umbrella is only about a fourth as great as that of the whole cassiopea. Table 17 records the measure ments on 3 cassiopeas and on their umbrellas after removal.
Since the respiration is influenced by the muscular activity or pulsa tion-rate, and the latter is not constant, it was decided to remove the ganglia (rhopalia) that induce the pulsations and start a continuous contraction-wave running around the subumbrella (the middle third of which has no neuro-muscular tissue, Mayer, 1908). The rhopalia were cut out by means of a cork-borer and the wave started by electrical stimulation. It was noticed, however, that the contraction-wave, apparently constant for short intervals of time, changed more rapidly at first and then more slowly, but never became absolutely constant, the change being perhaps associated with shrinkage of the umbrella. The effect of shortening and stretching on the contraction-wave was therefore studied. The rate of the contraction-wave depends on the rate of the nerve-impulse around the circuit of the nerve-muscle layer, but does not depend solely on the rate in the neuraxon, since there are numerous synapses, and furthermore, the path of the impulse is zigzag. Prof. L. R. Cary kindly showed me a stained preparation of the nervous network of the subumbrella. Concentric rings cut from the umbrella are capable of maintaining a trapped wave for some time, but if the ring is too narrow, the wave can not be started or soon ceases after being started. Trapped waves can be started in 2 or 3 concentric rings cut from the umbrella and the wave revolves about the inner ring more often per second than about the outer ring, but the revolutions per second are not in exact inverse proportion to the mean diameters of the rings or to the diameters of the inside tracks or holes in the rings. One subumbrella, 11.5 cm. in diameter, was cut into two rings and waves were trapped in them. The wave in the inner ring made 2.5 revolu tions per second and the wave in the outer ring made 2 revolutions per second.