The first specimen of Ammocoetes branchialis that showed histologically any atrophic changes in the endostyle was taken on July 16. These changes proceeded relatively rapidly for about a month, after which the endostyle as such was no longer recognized. All specimens examined after August 15 showed in cross section the characteristic ductless follicles more or less completely formed. More gradual and minor changes in the way of further absorption of cell remnants and completion of the follicles continued at least until September 1. Two specimens taken from the creek on September 4, 1911. showed complete follicle formation with some stainable colloid (figures 14 and 15). There was still yellow granular pigment in the fibrous tissue between the follicles. In two specimens taken on October 14, 1909, the pigment was absent and the follicles were more closely set, larger, and contained homogenous colloid. In the twenty-four specimens of ammocoetes studied, there were variations in the time of the onset of metamorphosis. There may also be variations in the rate of progress of the changes in different specimens. There is no evidence that removal of the animals from their native environment to the laboratory either increases or decreases the rate of metamorphosis. Schneider states that he was unable to get specimens kept in the laboratory to undergo metamorphosis. Gage, however, has repeatedly observed the metamorphosis under laboratory conditions, and the six of our specimens kept in the laboratory—some for forty days—remained in excellent condition and the metamorphosis proceeded as well as in those living in the creek. I know of no observations bearing on the question as to whether the metamorphosis may be hastened or delayed as it can be in tadpoles and other amphibia. It is probable, however, that physical conditions greatly influence the transformation.
These observations as to the length of time from the inception to the completion of metamorphosis indicate that a month and probably longer is necessary for the lake and brook lampreys of Central New York. This is in agreement with the observations of Gage and of Muller on metamorphosis in general, but is at variance with the views of Bujor, who states that the process takes place within three to four days.
The first endostylar changes are a gradual shrinkage in the whole organ with thickening of the capsule and septum and proliferation of the connective tissue in the periendostylar zone. The tongue anlage is developed in this thickening just dorsal to the endostyle and anterior to the gland orifice. The size of the chambers progressively decreases and with the thickening of the septum the halves of the endostyle are both absolutely and relatively more separated. All the five types of epithelia are affected, the first to show the change being type I, the four fan-shaped bundles of cuneiform cells of each half of the endostyle. These disappear totally quite early. The next type to show marked changes is type III, or the cells with yellow pigment granules. Here the change is progressive and these cell groups in different stages of atrophy may be traced through to the fully developed follicles. The epithelium of type V, or the endothelial-like lining of the parietal walls of the chambers, is piled up and extruded laterally as the chambers contract or shrink. These cells in different stages of atrophy may be followed until the metamorphosis is nearing completion. It is certain that the cells of types I, III, and V play no part in the formation of the ductless follicles. With types II and IV the question is not so easily settled as it is from one or the other or from both of these types that the permanent follicles arise. One can say definitely that type IV plays the major role, but whether the cells of type II after fusion with the basal group of type IV do not also share in the formation of the ventral follicle of the given chamber, I cannot decide, but from the evidence obtainable this seems probable. It is significant that the cells of type IV are continuous with, and indistinguishable from, the cells lining the orifice and are continued anteriorly in the deep pharyngeal groove and peripharyngeal grooves as well as posteriorally from the orifice in the small pharyngeal groove. As to the fate of this extraglandular epithelium of type IV I have no data save that with the closing of the orifice and the formation of the permanent branchial sac these grooves with their ciliated epithelium disappear and the whole sac comes to be lined with plain stratified epithelium. The fact that the cells of the pharyngeal grooves and the lining cells of the gland orifice are continuous with the cells of the endostyle from which the permanent thyroid follicles are formed is not without significance in relation to the development of the thyroid of the higher chordates. One or more very large follicles are formed from the lower portion of this orificial epithelium of type IV.
Four ductless follicles are the maximum number that may be formed primarily in each half of the endostyle from the four areas of epithelium of type IV. From the specimens studied this maximum is frequently not obtained. Posterior to the orifice where four chambers exist, each corresponding to one half of an anterior chamber, but two follicles may be formed from each chamber, but in the coil these are proportionately increased, in cross section. Most of the detailed studies here recorded have been made on the part of the endostyle posterior to the coil where the simplest conditions exist. Here two follicles are ordinarily formed from each chamber.
In cross sections the follicles are at first only long tubules whose cavities are the remnants of the original endostyle chambers, but when the metamorphosis is completed each of these primary tubules is cut up into several elongated closed sacs corresponding to the true ductless follicles of all higher chordates. New follicles also arise by budding from these primary ones, and this process is probably of normal occurrence at the metamorphosis.