Twelve normal monkeys inoculated on the mucous membranes of the nose or nose and mouth with a strain of Bacillus influenzæ; originally isolated in pure culture from the pleural exudate of a case of empyema following influenzal pneumonia in man and subsequently raised in virulence by animal passage, developed an acute self-limited respiratory disease of from 3 to 5 days duration, characterized by sudden onset with profound prostration, the development of rhinitis and tracheobronchitis, with sneezing, cough, and the outpouring of a scanty mucoid, or mucopurulent exudate, a variable febrile reaction, and either a leucopenia or no significant change in the leucocyte count. This disease was complicated in five instances by purulent sinusitis of one or both antra, in three by bronchopneumonia. Bacillus influenzæ was recovered at autopsy from the lesions of the disease either in pure culture or in association with organisms that are normal inhabitants of the upper respiratory tract of monkeys. Of ten normal monkeys injected intratracheally with the same strain of Bacillus influenzæ, seven developed bronchopneumonia, two developed tracheobronchitis without pneumonia, and one resisted infection. The general symptoms and duration of the disease were similar to those of the preceding group. There were a severe cough and accelerated respirations. Bacillus influenzæ was recovered in pure culture from the lungs, bronchi, or trachea in the animals killed during the active stage of the disease. It disappeared promptly from the respiratory tract with recovery. The significance of the first series of experiments in which monkeys were inoculated in the upper respiratory tract is twofold. First, they establish the fact that Bacillus influenzæ can initiate in monkeys an acute infection of the normal mucous membranes of the upper respiratory tract; that is, it can act as a primary incitant of respiratory infection without the assistance of a preceding or concomitant contributing cause. In this respect it differs radically from the pneumococcus and Streptococcus hæmolyticus, since experiments previously reported2, 4 have shown that neither of these organisms possesses the property of initiating an infection of the normal mucous membranes of the upper respiratory tract of monkeys, even though the strains used were incalculably more virulent for monkeys than the strain of Bacillus influenzæ used in the foregoing experiments. Secondly, the experiments show that Bacillus influenzæ infection of the mucous membranes of the upper respiratory tract may spread by continuity to the paranasal sinuses, setting up an acute sinusitis, that it spreads readily to the lower respiratory tract, producing a tracheobronchitis and permitting the ready invasion of secondary bacteria, and that it may penetrate as far as the terminal bronchioles, alveolar ducts, atria, and alveoli, there setting up a bronchiolitis and true bronchopneumonia. In these respects it likewise differs radically from the pneumococcus and Streptococcus hæmolyticus which do not possess these pathogenic properties as previous experiments have shown.2, 4 The bearing of these facts on the possible etiologic relation of Bacillus influenzæ to influenza is important, since they show that Bacillus influenzæ possesses certain definite primary pathogenic properties which distinguish it and therefore separate it from the group of recognized secondary organisms in influenzal complications, of which the pneumococcus and the streptococcus are the most frequent. The possible etiologic relation of Bacillus influenzæ to influenza is further supported by the character of the respiratory disease that occurred in the monkeys. The sudden onset with profound prostration, the absence of leucocytosis or often a leucopenia, the congestion of the mucous membranes of the respiratory tract, the development on the 2nd or 3rd day of an irritative cough due to an inflammatory tracheitis or tracheobronchitis, the brief self-limited course of the infection, and the irregular febrile reactions are all characteristic of influenza. Many of these symptoms were in striking contrast with the symptoms and course of pneumococcus or streptococcus infections in monkeys in which there were no prostration at onset, invariable leucocytosis, and infrequent cough developing only late in the disease. While all the above features of the disease produced in monkeys are characteristic of influenza in man, none are pathognomonic and, in fact, it is doubtful whether uncomplicated influenza possesses any pathognomonic features by which it may be diagnosed certainly in the absence of an epidemic. Even during epidemic times many respiratory infections arise which, though presumably influenza, it is impossible to diagnose as such with certainty. Nor does pathology help in this respect, since there would appear to be no established distinctive lesions of uncomplicated influenza in man, nor for that matter of the complications of influenza, apart from the complications which have been ascribed by Pfeiffer,5 MacCallum,6 Wolbach,7 and others to infection with Bacillus influenzæ because of the association of Bacillus influenzæ in pure culture with these complications. For these reasons, although the disease produced in monkeys appears to be essentially identical with influenza in man with respect to its clinical course and complications, it is impossible to determine certainly whether it is actually so. The experiments are advanced, therefore, as evidence in favor of the etiologic relation of Bacillus influenzæ to influenza, though they do not permit of a definite conclusion in this respect. Their bearing upon the relation of Bacillus influenzæ to certain of the complications of influenza would appear to be reasonably conclusive. The recovery of Bacillus influenzæ in pure culture at autopsy from the antra, from the trachea and bronchi, and from the lungs in some of the animals developing sinusitis, bronchiolitis, and a characteristic type of bronchopneumonia confirms by animal experiment the etiologic relation of Bacillus influenzæ to these complications of influenza, which hitherto has rested solely upon the frequent association of the influenza bacillus with these lesions in man. The production of tracheobronchitis and the same type of bronchopneumonia by the intratracheal injection of Bacillus influenzæ in the second series of experiments serves as additional confirmation of this, but has no direct bearing on the etiologic relation of Bacillus influenzæ to uncomplicated influenzæ.