PATHOLOGY AND MORBID ANATOMY.--The seat of the morbid process being the connective tissue, those localities of the larynx in which this tissue is most abundantly interposed between the mucous membrane and the cartilage are most liable to infiltration. I must say from my own experience that the epiglottis--particularly the glosso-epiglottic region--is most frequently affected,[3] next the ary-epiglottic folds, then the arytenoid region, and then the ventricular folds. The ventricles and the vocal bands are very rarely involved. Infraglottic oedema is still more rare, and is never an extension of the supraglottic. The disease is never a primary one, and, though seated in the submucous connective tissue, it may have started with inflammation of either the overlying mucous membrane or the underlying perichondrium. Effusion of blood is generally limited to traumatic cases, but has ensued from mercurialization, small-pox, and typhus; purulent infiltration and abscess formation is the result of phlegmonous inflammation and breaking down of the tissue, occurring especially in the cushion of the epiglottis and in the ventricular and ary-epiglottic folds; but as a rule the effusion in laryngeal oedema is of a serous or sero-purulent character.[4] In infraglottic oedema it is said to be fibrinous. [Footnote 3: According to Sestier, the ary-epiglottic folds are affected in nearly every case, either alone or together with other parts.] [Footnote 4: In 90 cases Sestier found the infiltration serous 60 times, sero-gelatinous 6, sero-purulent 9, sero-purulent with plastic lymph 4, purulent 8 times, sero-sanguineous twice, and sanguineous once.] The mucous membrane covering the oedematous structures is tense and discolored; except in very inflammatory conditions it is yellowish, shimmering, and pallid. On cutting into the diseased parts often but little exudation takes place, and sometimes even squeezing between the fingers does not suffice to cause disgorgement.[5] After the fluid is evacuated the parts collapse and the mucous membrane is left wrinkled and folded. [Footnote 5: In 23 autopsies Sestier found that incisions into the oedematous structures made the liquid run out either without any or with slight pressure 10 times; with repeated pressure, with difficulty and only in small quantity, 6 times; and not at all, in spite of repeated incisions and pressure, 7 times.]