In a series of observations upon the thyroids of brook trout, extending from the time of hatching to fish of four or more years old, we have been able to follow step by step the development of thyroid hyperplasia. In these hatchery trout, thyroid overgrowth may be recognized anatomically at the end of the second month of extra-oval life. Under the conditions in which these fish live, the thyroid overgrowth progressively increases, at first filling the aortic space and then invading the surrounding bone and muscle. In our series, this extension into the adjacent tissues was not recognizable until the fourth month of extra-oval life. The time when invasion of the surrounding tissues is first noticed is, of course, subject to great variation in different years and different hatcheries, depending on the general hygienic conditions, food, water supply, etc. If the growth continues, as it usually does, throughout the rapid growing period of the fish, it may first be detected clinically by a reddening of the pharyngeal floor as the thyroid tissue spreads to the submucosa. Later, definite external goitres appear ventrally. These may be present as early as the sixth month of life. The thyroid being, as in mammals, most active during the growing period, the greatest number of visible goitres appear during the second and third years of life. In older fish the thyroid again becomes less active and there is a tendency toward spontaneous recovery. There is no noteworthy change in the anatomical appearance of the thyroid growth, whether seen in very young or in older fish, other than that clearly dependent on the age of the fish, mechanical factors, as size of the goitre, and complicating factors, as infection, hemorrhage, degeneration, etc. There exists, therefore, no anatomical basis for the diagnosis of cancer in the older fish that is not also present in the youngest fish with thyroid hyperplasia. All the appearances of invasion, atypical cell growth, etc. are the results of the progressive growth and consequent extension of a non-encapsulated epithelial tissue along the paths of least resistance. In many of the larger goitres there are more or less distinct areas of thyroid hyperplasia, histologically different from the surrounding thyroid tissue, which we look upon as benign tumors comparable to the benign tumors seen in human goitres. We class them as tumors because they do not react with iodin as does the ordinary hyperplasia. To call these tumors cancer is going beyond our present knowledge and is unjustified, even though analogy with mammalian tumors suggests that certain ones might proceed to true carcinomata. The most common type of the benign tumors is the arborescent-papillomatous form and its many modifications. Infection is a frequent complication in the larger goitres, for the reasons given. It is necessary to recognize this factor in interpreting the histological conditions met with, as otherwise, using mammalian standards, some of the appearances observed could easily be mistaken for sarcoma. The effect of iodin on the thyroid hyperplasia has been studied in the early, middle, and late stages. All stages react with iodin, the mild degrees seemingly more rapidly than the severer degrees. Thus the early stages undergo involution in from two to three weeks, while the late stages may require one to two months. The process, therefore, is slower than the iodin reaction time in mammalian hyperplasia. True tumors do not react with iodin as does ordinary hyperplasia. Infection or other complication modifies the reaction. The reaction with iodin is a specific test for functional hyperplasia of the thyroid. As we have not been able to find any stage in the process that does not react with iodin, we must conclude that there is none that may be looked upon biologically as cancer. Just as iodin invariably stops the hyperplasia and causes the thyroid to return to the colloid or resting state (from which it may undergo hyperplasia a second, third, or more times, exactly as in mammals), so also spontaneous involution occurs when the fish are transferred to a natural environment. This spontaneous involution has been followed in experiments and has been seen in examples taken at random from the streams. We have never seen an exception to this rule, save in the case of true tumors. As pointed out above, recovery does not imply a disappearance of the goitre; it implies a cessation of growth and return of the active hyperplasia to the colloid or resting stage. Thyroid hyperplasia is a compensatory reaction, the exact cause of which is still to be sought. There is no evidence that in fish or in man it is either infectious or contagious. All the biological data at present available favor the view that fish goitre in common with mammalian goitre is the symptomatic manifestation of a metabolic and nutritional disturbance. There are three major conditions which, in some way still obscure, influence the thyroid growth; namely, a limited water supply, overcrowding, and overfeeding with a highly artificial and incomplete food. The water of the hatchery observed is not intrinsically goitre-producing, as the fish will not develop goitre unless at least the factor of overfeeding with an incomplete food operates at the same time. On the other hand, they recover if the overfeeding and overcrowding are corrected, although they remain in the same water. Therefore it seems probable that the food is the major factor acting to bring about a fault of nutrition favorable for goitre development. It is impossible at this time to suggest what elements in the food may be at fault, whether, for example, it is deficiency or excess or disproportion in the relative food values.