- Email: [email protected]
A method for monitoring end-tidal CO2 during neurophysiological experiments on small laboratory animals
Journal of Neuroscience Methods, 7 (1983) 227-233
227
Elsevier Biomedical Press
A method for monitoring end-tidal CO 2 during neurophysiological experiments on small laboratory animals Gregory R. Bock, Graeme K. Yates 1 and Padma A. Moorjani MRC Institute of Hearing Research, University of Nottingham, University Park, Nottingham NG7 2RD (U.K.) (Received July 5th, 1982) (Accepted August 8th, 1982)
Key words: end-tidal CO 2 --guinea pig--airway adaptor--Capnometer The Hewlett Packard Model 47210A Capnometer, intended for monitoring expired C O 2 in human patients, can be modified for use in acute neurophysiological studies on small animals. The airway adaptor which would normally be attached to an endotracheal tube is modified and its internal volume is reduced to 0.2 ml, and is incorporated into a guinea pig head-holder. We have verified that the modification does not alter the Capnometer's accuracy. There is no significant increase in resistance in the airway when the adaptor is in place, and the waveform of instantaneous CO 2 shows rapid changes on inspiration and expiration with well-resolved peaks and troughs. Measurement of arterial pCO 2 suggests that the Capnometer and modified airway adaptor form an accurate system for obtaining a continuous record of end-tidal CO 2 in small guinea pigs.
Introduction
Expired pCO 2 is closely related to blood CO2 level, which is in turn an index of general CNS function. Deterioration in end-tidal CO2 values accompanies depressed CNS function, changes in blood pH, changes in arterial blood flow, etc. It is important to detect such changes if they occur during an acute neurophysiological experiment, since general CNS depression could influence the parameter under study. For example, in studies of single cochlear nerve fibres it has been shown that poor sensitivity to sound stimuli and poor frequency selectivity are concomitants of a decline in the condition of the animal (Evans, 1974). Similar examples could be chosen from other areas of neurophysiological research. Since adequate ventilation is essential to the physiological well-being of any animal it follows that monitoring of
i Present address: Department of Physiology, University of Western Australia, Nedlands, WA 6009, Australia. 0165-0270/83/0000-0000/$03.00 © 1983, Elsevier Science Publishers
228 ventilation can help in assessing the reliability of results with many experimental procedures. Most commercial CO z analyzers are designed for use in human patient monitoring, and rely on the extraction of samples of expired gas into a measuring cell, where CO z tension is deduced from the infra-red absorption of the gas. The size of the sample cell is small relative to human respiratory tidal volume, but large relative to the tidal volume of small laboratory animals (e.g. 1-4 ml in the guinea pig, Green, 1979). Consequently, such devices do not give accurate respiratory waveforms or end-tidal CO 2 readings when used with small animals. In our laboratory we wished to monitor end-tidal CO z tension during acute studies of cochlear physiology in guinea pigs in the weight range 150-500 g. The object of the present report is to describe a simple modification to the HewlettPackard Model 47210A Capnometer which renders this device suitable for such monitoring.
Modification of the Capnometer. The Hewlett-Packard Model 47210A Capnometer measures infra-red absorption directly in the airway. An airway adaptor (Fig. 1) fits onto an endotracheal tube, and a detector head clamps over this adaptor. The detector head, which is connected by a cable to the main instrument, shines an infra-red light through two sapphire windows in the airway adaptor and measures the resulting absorption. CO z is then calculated and displayed on the main instrument. Our modification was designed to reduce the volume of the airway adaptor without obstructing the infra-red light pathway between the two sapphire windows. Since the machine can measure CO 2 tension only between the windows, it follows
Fig. 1. Left: drawing of the unmodified airway adaptor. Right: drawing showingthe cut airway adaptor (bottom) and the perspex insert (top). Positions of holes in the perspex are indicated. In use, the piece of perspex is pushed into the groovein the airway adaptor so that the large transverse hole coincides with the sapphire windows.
229
that such a modification should not alter the accuracy of the device if no modification is made to the instrument itself. The airway adaptor consists of a short T-section die-cast tube, with the sapphire windows mounted across the stem of the T (Fig. 1). The head of the T-section and the ends were carefully cut off with a handsaw, leaving only the section carrying the windows (Fig. 1). Next, a piece of perspex 31 mm x 12 mm x 3 mm is prepared. This will fit into the space between the windows and fill it completely. A hole 6 mm in diameter is drilled so that the space between the sapphire windows remains unobstructed, and three longitudinal holes (2.2 mm diameter) are drilled to complete the air pathway (Fig. 1). This piece of perspex is then fitted tightly into the groove in the airway adaptor. The perspex can be removed at any time if a different configuration of the airway holes is required. The calculated dead space (i.e. the total volume of the air pathways in the modified adaptor) of the adaptor illustrated in Fig. 1 is 0.2 ml, but this is reduced considerably when cannula tubing is inserted to make the airway connections. In use, the modified airway adaptor is mounted onto a standard guinea pig head-holder (Fig. 2) so that the end of the adaptor containing only one opening is as close as possible to the animal. Animals are anaesthetized with the neuroleptanalgesia technique of Evans (1978) and a tracheal cannula (35 mm of 2.2 mm o.d. polythene tubing) is inserted. When the anaesthetized animal is clamped into the head-holder the free end of the tracheal cannula is inserted into the hole in the modified airway adaptor. If the animal is not to be artificially ventilated, the two holes at the other end of the airway adaptor are left unobstructed. Alternatively, if
Fig. 2. The photograph shows a guinea pig head-holder, turned upside down to show the airway adaptor in position. The detector head, which clamps over the airway adaptor, is shown separately in the foreground.
230 the animal is to be artificially respired, two tubes are inserted into these openings and connected to the inlet and outlet tubes of a small animal ventilator (CF Palmer Model 5255). Finally, the head-holder can be rotated to any position appropriate for surgery and the detector head is fitted to the modified adaptor. No modification is made either to the capnometer itself or to the detector head.
Validation We first verified that our modification did not interfere with the operation of the Capnometer. This was achieved by mixing in various proportions the output of 3 gas cylinders, one containing an O2/CO 2 mixture (95%/5%), the second containing 100% 02, and the third containing 100% N20. The resulting mixture was then passed through the modified adaptor and an unmodified adaptor in series with it in the same gas pathway. With this system we could alter the partial pressure of CO 2 from 0 mm Hg to 38 mm Hg. For any fixed CO 2 concentration within this range, the detector head always gave the same reading when clamped to the modified or unmodified adaptors. The Capnometer has front-panel switchable compensation for high O: or the presence of N20, and these compensations were unaffected by the modifications to the airway adaptor. Given that the modification did not alter the Capnometer's reading, we next evaluated the effect of the adaptor on airway resistance. When a free-breathing guinea pig is connected to the modified airway adaptor there is no detectable change in respiration rate, and no sign that any respiratory distress has occurred. To obtain a more objective measure of the resistance of the modified airway adaptor, air pressure was measured directly in 4 animals using a probe consisting of a length of 0.6 mm diameter polythene tubing attached to a silicon pressure transducer (National Semiconductor type LX1601GBZ). Before the tracheal cannula was connected to the airway adaptor, the tip of the probe was inserted into the end of the cannula and the output of the pressure transducer was monitored on a chart recorder. The probe was then withdrawn and the tracheal cannula was inserted into the modified airway adaptor. The probe was then pushed through the airway adaptor until the tip was again in the entrance to the tracheal cannula, and the pressure was recorded on the chart recorder. Pressure fluctuations in the tracheal cannula were approximately 0.8 cm H20 peak-to-peak, and the presence of the airway adaptor made no detectable difference to these fluctuations. The Capnometer computes respiration rate and the displayed value is accurate with both free-breathing and artificially-respired guinea pigs provided there are no respiration abnormalities such as double breathing. When such abnormalities do occur, the Capnometer displays a 'respiration artefact' error code. The other pre-set alarm and error diagnostics are unaffected by the modification. The technical specification for the Capnometer states that the response time for an instantaneous change in CO 2 tension is less than 200 ms to 90% full reading, but we have not verified this directly. However, the measured instantaneous CO z tension is available as an analogue voltage output from the instrument and Fig. 3 shows a
231
~' 3O E E 20
I
I
1
t
I
J
I
2
)
~
I
4
I
I
I
5
1
6
Time (secs~ Fig. 3. The trace shows instantaneous pCO 2 recorded from artificially-ventilated guinea pig, The animal was deliberately hyperventilated to produce a high respiration rate.
sample chart recording of respiration waveform for an artificially respired guinea pig. The animal was deliberately hyperventilated in order to produce a high respiration rate, since a high respiration rate provided a more stringent test of the Capnometer's response. It can be seen that, even with a high respiration rate, the CO 2 level changes rapidly during inspiration and expiration, the peaks and troughs of the waveform are flat and well resolved, and the CO 2 tension recorded during inspiration is close to zero. Finally, we have verified the accuracy of the entire measurement system by comparing end-tidal CO 2 tension with blood CO 2 tension. Six guinea pigs were anaesthetized and placed in the head-holder. The carotid artery was cannulated and approximately 0.4 ml of blood extracted, and at the same time the indicated end-tidal CO 2 was noted. The samples were analyzed within 10 rain of extraction on a Corning model 175 blood gas analyzer, and the measured CO 2 in blood was
TABLE 1 E N D - T I D A L CO 2 R E A D I N G S A N D C O R R E S P O N D I N G A R T E R I A L pCO 2 VALUES IN 6 G U I N E A PIGS Animal number 6 was hyperventilated and number 4 hypoventilated in order to test at extremes of CO 2 tensions. Animal number
End-tidal pCO 2
Blood pCO 2
1
44 45 47 71 49 31
43 46 50 73 49 36
2 3 4 5 6
232 compared with the end-tidal C O 2 registered by the Capnometer. The results are shown in Table I. It can be seen that there was close agreement between the Capnometer's reading and the blood CO 2 measure. The detector head which produces the infra-red light contains a mechanical rotor to 'chop' the light beam. This produces a slight but detectable low-frequency vibration. We have measured this vibration using an opto-electronic displacement transducer (Yates, 1982). When the CO 2 detector head is resting on an elastic supporting surface (20 mm thick foam plastic) the detector head vibrates with an almost pure sinusoidal motion at approximately 40 Hz and with an amplitude of 12 /~m peak-to-peak. However, when the CO 2 detector head is clipped to the rigidly mounted animal head-holder we could detect no vibration of the head-holder above a noise floor of 0.1 /~m peak-to-peak.
Discussion The modification described here was intended primarily for use with artificially respired guinea pigs. However, it is unlikely that the exact configuration of the airway holes would be critical, so that various other arrangements could be chosen to suit other experimental procedures. Our observations suggest that connecting the modified airway adaptor to a guinea pig's tracheal cannula does not impede the animal's breathing. Small pressure fluctuations could be measured in the tracheal cannulas of anaesthetized guinea pigs, but these were not affected by the presence of the modified airway adaptor. The respiratory tidal volume of a guinea pig is in the range 1-4 ml (Green, 1979) and the measured dead space of the modified adaptor appears reasonable in relation to the animal's respiratory volume. In particular, it is less than the total volume of the guinea pig's mouth and nasal passages, which are removed from the animal's dead space by tracheotomy. The recordings of instantaneous CO 2 also suggest that the additional dead space is acceptable since the inspiration and expiration phases show rapid equilibration. We have taken two different approaches to validating our modifications to the Capnometer. The first assumes that the unmodified instrument is completely accurate and seeks only to show that the modification does not affect the behaviour of the device. The second approach is to look for some independent measure of the overall accuracy of the measuring system. Most of our observations relate to the first approach, and suggest that the modification does not alter any aspect of the Capnometer's performance. However, as an overall check on the absolute accuracy of the Capnometer and modified adaptor as a measurement system, we measured arterial CO 2 tension. Since there is no significant alveolar-arterial gradient for CO 2, we would expect measured end-tidal CO 2 tension to be close to arterial CO: tension, and our results show good agreement between the two measures. We therefore conclude that our modification of the Capnometer provides a reliable and efficient means of monitoring expired CO 2 in a guinea pig. Similar
233 modifications, involving less drastic reduction in the dead space, could readily be made for any animal larger than a guinea pig. Vibration produced by the detector head could constitute a potential problem with the system described here, but our measurements suggest that any vibrations transferred to the head-holder are less than 0.1 ~m in amplitude. Such vibration has not affected our ability to record extracellularly from neurones in the spiral ganglion, but may possibly be more problematic in experiments requiring intracellular techniques.
Acknowledgements The authors are indebted to D. Evans for his assistance with blood gas measurements, and to M.P. Haggard and P.H. Fentem for critical readings of the manuscript. Prof. Fentem's advice during the course of this work was greatly appreciated. This research was carried out within the framework of the European Economic Communities Concerted Action Programme on Hearing Impairment.
References Evans, E.F. (1974) Auditory frequencyselectivityand the cochlear nerve. In E. Zwicker and E. Terhardt (Eds.), Facts and Models in Hearing, Springer,New York, pp. 118-129. Evans, E.F. (1978) Neuroleptanaesthesiafor the guinea pig: an ideal anaestheticprocedure for physiological studies of the guinea pig, Arch. Otolaryngol., 105: 185-186. Green, C.J. (1979) Animal anaesthesia, Laboratory Animals, London. Yates, G.K. (1982) A sensitiveoptoelectronicdisplacement transducer for the neurophysiologicallaboratory, J. neurosci. Meth., 6: 103-11I.
227
Elsevier Biomedical Press
A method for monitoring end-tidal CO 2 during neurophysiological experiments on small laboratory animals Gregory R. Bock, Graeme K. Yates 1 and Padma A. Moorjani MRC Institute of Hearing Research, University of Nottingham, University Park, Nottingham NG7 2RD (U.K.) (Received July 5th, 1982) (Accepted August 8th, 1982)
Key words: end-tidal CO 2 --guinea pig--airway adaptor--Capnometer The Hewlett Packard Model 47210A Capnometer, intended for monitoring expired C O 2 in human patients, can be modified for use in acute neurophysiological studies on small animals. The airway adaptor which would normally be attached to an endotracheal tube is modified and its internal volume is reduced to 0.2 ml, and is incorporated into a guinea pig head-holder. We have verified that the modification does not alter the Capnometer's accuracy. There is no significant increase in resistance in the airway when the adaptor is in place, and the waveform of instantaneous CO 2 shows rapid changes on inspiration and expiration with well-resolved peaks and troughs. Measurement of arterial pCO 2 suggests that the Capnometer and modified airway adaptor form an accurate system for obtaining a continuous record of end-tidal CO 2 in small guinea pigs.
Introduction
Expired pCO 2 is closely related to blood CO2 level, which is in turn an index of general CNS function. Deterioration in end-tidal CO2 values accompanies depressed CNS function, changes in blood pH, changes in arterial blood flow, etc. It is important to detect such changes if they occur during an acute neurophysiological experiment, since general CNS depression could influence the parameter under study. For example, in studies of single cochlear nerve fibres it has been shown that poor sensitivity to sound stimuli and poor frequency selectivity are concomitants of a decline in the condition of the animal (Evans, 1974). Similar examples could be chosen from other areas of neurophysiological research. Since adequate ventilation is essential to the physiological well-being of any animal it follows that monitoring of
i Present address: Department of Physiology, University of Western Australia, Nedlands, WA 6009, Australia. 0165-0270/83/0000-0000/$03.00 © 1983, Elsevier Science Publishers
228 ventilation can help in assessing the reliability of results with many experimental procedures. Most commercial CO z analyzers are designed for use in human patient monitoring, and rely on the extraction of samples of expired gas into a measuring cell, where CO z tension is deduced from the infra-red absorption of the gas. The size of the sample cell is small relative to human respiratory tidal volume, but large relative to the tidal volume of small laboratory animals (e.g. 1-4 ml in the guinea pig, Green, 1979). Consequently, such devices do not give accurate respiratory waveforms or end-tidal CO 2 readings when used with small animals. In our laboratory we wished to monitor end-tidal CO z tension during acute studies of cochlear physiology in guinea pigs in the weight range 150-500 g. The object of the present report is to describe a simple modification to the HewlettPackard Model 47210A Capnometer which renders this device suitable for such monitoring.
Modification of the Capnometer. The Hewlett-Packard Model 47210A Capnometer measures infra-red absorption directly in the airway. An airway adaptor (Fig. 1) fits onto an endotracheal tube, and a detector head clamps over this adaptor. The detector head, which is connected by a cable to the main instrument, shines an infra-red light through two sapphire windows in the airway adaptor and measures the resulting absorption. CO z is then calculated and displayed on the main instrument. Our modification was designed to reduce the volume of the airway adaptor without obstructing the infra-red light pathway between the two sapphire windows. Since the machine can measure CO 2 tension only between the windows, it follows
Fig. 1. Left: drawing of the unmodified airway adaptor. Right: drawing showingthe cut airway adaptor (bottom) and the perspex insert (top). Positions of holes in the perspex are indicated. In use, the piece of perspex is pushed into the groovein the airway adaptor so that the large transverse hole coincides with the sapphire windows.
229
that such a modification should not alter the accuracy of the device if no modification is made to the instrument itself. The airway adaptor consists of a short T-section die-cast tube, with the sapphire windows mounted across the stem of the T (Fig. 1). The head of the T-section and the ends were carefully cut off with a handsaw, leaving only the section carrying the windows (Fig. 1). Next, a piece of perspex 31 mm x 12 mm x 3 mm is prepared. This will fit into the space between the windows and fill it completely. A hole 6 mm in diameter is drilled so that the space between the sapphire windows remains unobstructed, and three longitudinal holes (2.2 mm diameter) are drilled to complete the air pathway (Fig. 1). This piece of perspex is then fitted tightly into the groove in the airway adaptor. The perspex can be removed at any time if a different configuration of the airway holes is required. The calculated dead space (i.e. the total volume of the air pathways in the modified adaptor) of the adaptor illustrated in Fig. 1 is 0.2 ml, but this is reduced considerably when cannula tubing is inserted to make the airway connections. In use, the modified airway adaptor is mounted onto a standard guinea pig head-holder (Fig. 2) so that the end of the adaptor containing only one opening is as close as possible to the animal. Animals are anaesthetized with the neuroleptanalgesia technique of Evans (1978) and a tracheal cannula (35 mm of 2.2 mm o.d. polythene tubing) is inserted. When the anaesthetized animal is clamped into the head-holder the free end of the tracheal cannula is inserted into the hole in the modified airway adaptor. If the animal is not to be artificially ventilated, the two holes at the other end of the airway adaptor are left unobstructed. Alternatively, if
Fig. 2. The photograph shows a guinea pig head-holder, turned upside down to show the airway adaptor in position. The detector head, which clamps over the airway adaptor, is shown separately in the foreground.
230 the animal is to be artificially respired, two tubes are inserted into these openings and connected to the inlet and outlet tubes of a small animal ventilator (CF Palmer Model 5255). Finally, the head-holder can be rotated to any position appropriate for surgery and the detector head is fitted to the modified adaptor. No modification is made either to the capnometer itself or to the detector head.
Validation We first verified that our modification did not interfere with the operation of the Capnometer. This was achieved by mixing in various proportions the output of 3 gas cylinders, one containing an O2/CO 2 mixture (95%/5%), the second containing 100% 02, and the third containing 100% N20. The resulting mixture was then passed through the modified adaptor and an unmodified adaptor in series with it in the same gas pathway. With this system we could alter the partial pressure of CO 2 from 0 mm Hg to 38 mm Hg. For any fixed CO 2 concentration within this range, the detector head always gave the same reading when clamped to the modified or unmodified adaptors. The Capnometer has front-panel switchable compensation for high O: or the presence of N20, and these compensations were unaffected by the modifications to the airway adaptor. Given that the modification did not alter the Capnometer's reading, we next evaluated the effect of the adaptor on airway resistance. When a free-breathing guinea pig is connected to the modified airway adaptor there is no detectable change in respiration rate, and no sign that any respiratory distress has occurred. To obtain a more objective measure of the resistance of the modified airway adaptor, air pressure was measured directly in 4 animals using a probe consisting of a length of 0.6 mm diameter polythene tubing attached to a silicon pressure transducer (National Semiconductor type LX1601GBZ). Before the tracheal cannula was connected to the airway adaptor, the tip of the probe was inserted into the end of the cannula and the output of the pressure transducer was monitored on a chart recorder. The probe was then withdrawn and the tracheal cannula was inserted into the modified airway adaptor. The probe was then pushed through the airway adaptor until the tip was again in the entrance to the tracheal cannula, and the pressure was recorded on the chart recorder. Pressure fluctuations in the tracheal cannula were approximately 0.8 cm H20 peak-to-peak, and the presence of the airway adaptor made no detectable difference to these fluctuations. The Capnometer computes respiration rate and the displayed value is accurate with both free-breathing and artificially-respired guinea pigs provided there are no respiration abnormalities such as double breathing. When such abnormalities do occur, the Capnometer displays a 'respiration artefact' error code. The other pre-set alarm and error diagnostics are unaffected by the modification. The technical specification for the Capnometer states that the response time for an instantaneous change in CO 2 tension is less than 200 ms to 90% full reading, but we have not verified this directly. However, the measured instantaneous CO z tension is available as an analogue voltage output from the instrument and Fig. 3 shows a
231
~' 3O E E 20
I
I
1
t
I
J
I
2
)
~
I
4
I
I
I
5
1
6
Time (secs~ Fig. 3. The trace shows instantaneous pCO 2 recorded from artificially-ventilated guinea pig, The animal was deliberately hyperventilated to produce a high respiration rate.
sample chart recording of respiration waveform for an artificially respired guinea pig. The animal was deliberately hyperventilated in order to produce a high respiration rate, since a high respiration rate provided a more stringent test of the Capnometer's response. It can be seen that, even with a high respiration rate, the CO 2 level changes rapidly during inspiration and expiration, the peaks and troughs of the waveform are flat and well resolved, and the CO 2 tension recorded during inspiration is close to zero. Finally, we have verified the accuracy of the entire measurement system by comparing end-tidal CO 2 tension with blood CO 2 tension. Six guinea pigs were anaesthetized and placed in the head-holder. The carotid artery was cannulated and approximately 0.4 ml of blood extracted, and at the same time the indicated end-tidal CO 2 was noted. The samples were analyzed within 10 rain of extraction on a Corning model 175 blood gas analyzer, and the measured CO 2 in blood was
TABLE 1 E N D - T I D A L CO 2 R E A D I N G S A N D C O R R E S P O N D I N G A R T E R I A L pCO 2 VALUES IN 6 G U I N E A PIGS Animal number 6 was hyperventilated and number 4 hypoventilated in order to test at extremes of CO 2 tensions. Animal number
End-tidal pCO 2
Blood pCO 2
1
44 45 47 71 49 31
43 46 50 73 49 36
2 3 4 5 6
232 compared with the end-tidal C O 2 registered by the Capnometer. The results are shown in Table I. It can be seen that there was close agreement between the Capnometer's reading and the blood CO 2 measure. The detector head which produces the infra-red light contains a mechanical rotor to 'chop' the light beam. This produces a slight but detectable low-frequency vibration. We have measured this vibration using an opto-electronic displacement transducer (Yates, 1982). When the CO 2 detector head is resting on an elastic supporting surface (20 mm thick foam plastic) the detector head vibrates with an almost pure sinusoidal motion at approximately 40 Hz and with an amplitude of 12 /~m peak-to-peak. However, when the CO 2 detector head is clipped to the rigidly mounted animal head-holder we could detect no vibration of the head-holder above a noise floor of 0.1 /~m peak-to-peak.
Discussion The modification described here was intended primarily for use with artificially respired guinea pigs. However, it is unlikely that the exact configuration of the airway holes would be critical, so that various other arrangements could be chosen to suit other experimental procedures. Our observations suggest that connecting the modified airway adaptor to a guinea pig's tracheal cannula does not impede the animal's breathing. Small pressure fluctuations could be measured in the tracheal cannulas of anaesthetized guinea pigs, but these were not affected by the presence of the modified airway adaptor. The respiratory tidal volume of a guinea pig is in the range 1-4 ml (Green, 1979) and the measured dead space of the modified adaptor appears reasonable in relation to the animal's respiratory volume. In particular, it is less than the total volume of the guinea pig's mouth and nasal passages, which are removed from the animal's dead space by tracheotomy. The recordings of instantaneous CO 2 also suggest that the additional dead space is acceptable since the inspiration and expiration phases show rapid equilibration. We have taken two different approaches to validating our modifications to the Capnometer. The first assumes that the unmodified instrument is completely accurate and seeks only to show that the modification does not affect the behaviour of the device. The second approach is to look for some independent measure of the overall accuracy of the measuring system. Most of our observations relate to the first approach, and suggest that the modification does not alter any aspect of the Capnometer's performance. However, as an overall check on the absolute accuracy of the Capnometer and modified adaptor as a measurement system, we measured arterial CO 2 tension. Since there is no significant alveolar-arterial gradient for CO 2, we would expect measured end-tidal CO 2 tension to be close to arterial CO: tension, and our results show good agreement between the two measures. We therefore conclude that our modification of the Capnometer provides a reliable and efficient means of monitoring expired CO 2 in a guinea pig. Similar
233 modifications, involving less drastic reduction in the dead space, could readily be made for any animal larger than a guinea pig. Vibration produced by the detector head could constitute a potential problem with the system described here, but our measurements suggest that any vibrations transferred to the head-holder are less than 0.1 ~m in amplitude. Such vibration has not affected our ability to record extracellularly from neurones in the spiral ganglion, but may possibly be more problematic in experiments requiring intracellular techniques.
Acknowledgements The authors are indebted to D. Evans for his assistance with blood gas measurements, and to M.P. Haggard and P.H. Fentem for critical readings of the manuscript. Prof. Fentem's advice during the course of this work was greatly appreciated. This research was carried out within the framework of the European Economic Communities Concerted Action Programme on Hearing Impairment.
References Evans, E.F. (1974) Auditory frequencyselectivityand the cochlear nerve. In E. Zwicker and E. Terhardt (Eds.), Facts and Models in Hearing, Springer,New York, pp. 118-129. Evans, E.F. (1978) Neuroleptanaesthesiafor the guinea pig: an ideal anaestheticprocedure for physiological studies of the guinea pig, Arch. Otolaryngol., 105: 185-186. Green, C.J. (1979) Animal anaesthesia, Laboratory Animals, London. Yates, G.K. (1982) A sensitiveoptoelectronicdisplacement transducer for the neurophysiologicallaboratory, J. neurosci. Meth., 6: 103-11I.