Primary immunisations in Liverpool.

small cannula placed in each of the anterior nares. The carbon dioxide profile was sampled continuously by a capnograph similar to that used by Bolton et al. The infant was, at the time of the recording, allowed to assume his own head position after being placed to sleep prone on a soft, bark filled mattress similar to the ones described by Ponsonby et al 2 and reported to be associated with an increased relative risk of sudden infant death syndrome (SIDS). The infant's head was also completely covered with a spun polyester filled quilt, the door was closed, and the ventilation grills were taped over. These circumstances were designed to promote 'rebreathing' of expired gases. The figure illustrates the breath by breath concentration of carbon dioxide at the nose and is typical of many other investigations that we have performed.3 This section of the trace was extracted some three minutes after commencement of recording and extends for a further two minutes. It is clear that some accumulation of carbon dioxide does occur but this is at much lower concentrations than those recorded from either mechanical or mathematical models. Furthermore, and most significantly, there is no observed tendency for the carbon dioxide to continue to increase; rather a 'steady state' is reached at a new carbon dioxide concentration. Figure 2 in their paper may afford some clue to the apparent discrepancy between the inexorable increase in carbon dioxide demonstrated by Bolton et al and the low level hyper-capnia which we see in real life. It can be seen from the carbon dioxide profile that in the model, end tidal as well as end inspired carbon dioxide rises. In fact they rise in parallel. The model is unphysiological in that it takes no account of the equilibration between lung carbon dioxide concentrations and mixed venous tension. The model also takes no account of body carbon dioxide stores (120 1 in an adult) which are large compared to lung stores and act as a compart-mentalised buffer. The model also takes no account of the extraction of oxygen from inspired air, the addition of 35 ml/min of carbon dioxide effectively creating a net outward pressure which will alter the dynamics of gas mixing in the bedding. The upper panel of our figure shows the carbon dioxide profile of a ventilatory response test carried out by the Read rebreathing method. In this test, modified by …