546 audible and visual alarm is given. Subsequent re-start of breathing stops the audible alarm, but leaves the visual alarm flashing until it is reset by the nursing staff.
We have used the device with babies varying in weight from 990 g. to 2500 g. and found it very reliable. Babies were sometimes clothed, sometimes naked, and often in incubators at ambient temperatures greater than 35 °C, and in each condition the performance was satisfactory. Like the Lewin air mattress, no skin attachments to the baby are necessary. However, the air mattress was not sensitive enough to record the respiratory movements of the 990 g. baby, whereas the new device still had reserves of sensitivity. The air mattress sometimes deflates from leakage or accidental puncture, and the sensing thermistor manifold is slightly cumbersome and rather easily broken. The new mattress was more robust; we also found that it was easy to clean, it was large enough for the bigger babies, and babies did not tend to roll off. On one occasion, when it was being tested on a very small baby, an impedance pneumograph apnoea monitor was also attached; in both instruments the sensitivity was high. When the baby had an apnoeic episode, the impedance apnoea alarm failed to signal an apnoeic episode because the impedance change of heart-beat was insufficient to activate the instrument, whereas the conductive mattress correctly sounded an alarm. When such a mattress at room temperature is placed in " a warm incubator it seems to need a 10-minute warming up " period, before the sensitivity is finally adjusted to requirements of monitoring the individual baby. Also when a baby is changed from, say, the lateral to the prone position slight adjustment of the sensitivity may be We found, however, that these adjustments necessary. were very easily made. This device does not, of course, give quantitative information of the depth or pattern of respiration. None the less, it performs its task of identifying episodes of apnoea and is sturdy, safe, reliable, unobtrusive, and easy to clean. We are advised that the instrument will probably cost under 100. Hammersmith Hospital, Du Cane Road, London W.12.
J. E. SMITH J. W. SCOPES.
IgA DEFICIENCY SIR,-Selective IgA deficiency has been associated with several diseases and has been reported also as an isolated defect unassociated with clinical abnormalities.1 The relationship of the IgA deficiency to the evolution of the patient’s clinical status is uncertain. However, only a few reports have considered the humoral immune response in the patients. Amman and Hong,1 reviewing 30 patients with IgA levels less than 0-05 mg. per ml., found humoral antibody formation to be quantitatively normal, and included normal humoral-antibody production as a criterion for the diagnosis of selective IgA deficiency. More recently, Nell et al. studied 13 patients from five families and found humoral-antibody response present in all patients tested. On the other hand, Hobbs,3 in reporting 24 patients with IgA levels under 50 mg. per 100 ml. (less than 20% of mean adult normal), found antibody deficiency in 7 out of 7 patients examined. In these three investigations, however, the responses of the individual classes of antibody were not examined. We have reported4a family in which 5 members from two generations have IgA levels not detectable by radial Amman, A. J., Hong, R. Medicine, 1971, 50, 223. Nell, P. A., Amman, A. J., Hong, R., Stiehm, E. R. Pediatrics, 1972, 49, 71. 3. Hobbs, J. R. Lancet, 1968, i, 110. 4. Tomkin, G. H., Mawhinney, H., Nevin, N. C. ibid. 1971, ii, 124. 1. 2.
IgG AND IgA STAINING AT1- DILUTION OF SERA
PATIENTS WITH SELECTIVE
immunodiffusion. This family has been investigated further by examining random serum-samples by indirect immunofluorescence5 for the presence of IgG and IgA antibody specific to poliovirus type 11, measles virus, and herpessimplex virus. The results are shown in the table. Similar investigations in the normal population have shown that positive IgG staining for these viruses is almost invariably
accompanied by IgA staining.5 Our findings demonstrate
characterised by diminution in the specific IgA response. It would be of interest to know if a similar absence of the specific response would be observed in these patients after stimulation with an antigen such as oral Dolio vaccine. HELEN MAWHINNEY MARGARET HAIRE.
Department of Microbiology, Queen’s University of Belfast. Royal Victoria Hospital, Belfast. Human Genetics
G. H. TOMKIN.
Department of Medical Statistics, Queen’s University of Belfast, Northern Ireland.
N. C. NEVIN.
MITRAL STENOSIS AND THE CRESCENDO PRESYSTOLIC MURMUR
SIR,-From their observations on patients with mitral stenosis, Criley and his colleagues postulated 6,7 that crescendo vibrations occurring after the onset of leftventricular contraction and during mitral valve closure result from an increasing velocity, although decreasing volume, of flow through a progressively diminishing valve orifice. In a similar study s on 23 patients with mitral stencsis and atrial fibrillation, we independently reached identical conclusions. However, neither Criley and coworkers nor your editorial commenting on their reports (May 20, p. 1107) clarified or correctly specified the point in "
time at which the mitral valve "closes". The exact " closure " point is important, since it marks the onset of the isovolumetric phase of left-ventricular systole and must take place 9,10 at the time of cross-over (P.c.o.) of leftventricular and left-atrial pressures. The major left-sided component (MI) of the first sound has been shown5 to occur after P.c.o., thus after valve closure but during the isovolumetric contraction phase, and is probably produced by maximal tension on the coapted leaflets, and their chordae, at the peak of their ascent into the left-atrial cavity. The interval between the onset of pressure-rise in the ventricle and the point P.c.o. is the pre-isovolumetric left-ventricular-contraction phase and, because of the high left-atrial pressure, is considerably prolonged in mitral stenosis. It is throughout this pre-isovolumetric phase of leftMawhinney, H., Haire, M., Adair, B. McC. Unpublished. 6. Criley, J. M., Feldman, I. M., Meredith, T. Am. J. Med. 1971, 51, 456. 7. Criley, J. M., Hermer, A. J. New Engl. J. Med. 1971, 285, 1284. 8. Lakier, J. B., Pocock, W. A., Gale, G. E., Barlow, J. B. Br. Heart J.
(in the press). Lakier, J. B., Fritz, V. U., Pocock, W. A., Barlow, J. B. S. Afr. J. med. Sci. 1970, 35, 85. 10. Lakier, J. B., Fritz, V. U., Pocock, W. A., Barlow, J. B. Br. Heart J. 1972, 34, 160. 9.