Maximum expiratory flows at 50% of vital capacity while breathing air and a He-O2 mixture and delta Vmax50 observed during three sittings (morning, noon and morning the day after) in reference subjects and in patients with bronchial hyperreactivity were submitted to statistical analysis by between sittings comparison, looking for correlation coefficients, paired "t" test and variation coefficient. Results of paired "t" test ruled out the influence of fatigue and of training in performing forced expiratory manoeuvres. Correlation coefficients relevant to absolute flows are highly significant in both groups, whereas coefficients of variation are satisfactory in reference subjects and slightly higher although still acceptable in patients with bronchial hyperreactivity. As far as delta Vmax50 is concerned, the results of statistical analysis are positively unacceptable, thus demonstrating that this arithmetic manipulation of absolute flows is the ground for a poor reproducibility of results.

Blood gas data for HbO2% and PaCO2 and determination of acid-base balance enable the biochemical features of bronchial obstruction to be related to primarily "intrinsic" or "extrinsic" factors. Long-term gas studies and ventilation mechanics data (lung elastic recoil pressure, maximum expiratory flow at different volumes and the corresponding recoil pressures) are also of assistance in this respect. Each type of bronchial obstruction thus has its own peculiarities, and these are also reflected in the level of metabolic compensation for respiratory acidosis.

Circumstances under which the use of oxygen-therapy in lung disease can be effective and harmless, depend upon a careful evaluation of its indications: they are suggested by the clinical need of correction of hypoxaemia as well as by the awareness of factors determining respiratory failure and of problems concerning O(2) transport and supply to tissues in health and disease. Blood gases monitoring enables to control the effects of treatment on arterial O2 and CO2 tensions thus giving all the useful data for oxygen administering particularly as far as components of hyperoxygenated mixtures, flow rate, duration, use of very effective low-risk devices (Venturi masks) are concerned. Correction of hypoxaemia involves the reduction of hypertension of the pulmonary circulation and hyperglobulia, improvement of tolerance of exertion, and attention to the metabolic compensation of respiratory acidosis. These results are influenced by the nature of the pathogenetic factors behind broncho-obstructive disease, which may lead to either a primarily "bronchitis" or a primarily "emphysematous" syndrome. An interesting feature relates to prognosis in the case of patients making home use of hyperoxygenated mixtures as part of a rehabilitation program, or to improve their quality of life. The cost and benifits of such treatment should be carefully weighed. Lastly, in the event of protracted treatment, attention must be paid to the possibility of toxicity and the means to be adopted for its prevention.