specific innervation of the muscle, so that when the nerves of a fast and of a

Quart. J. exp. Phy8iol. (1967) 52, 293-304 THE DIFFERENTIATION OF CONDUCTION VELOCITIES OF SLOW TWITCH AND FAST TWITCH MUSCLE MOTOR INNERVATIONS IN KITTENS AND CATS. By R. M. A. P. RIDGE.* From the Physiology Department, King's College, London. (Received for publication 8th June 1966) The difference in conduction velocity (maximum and mean) of the a motor nerve fibres to fast (flexor hallucis longus) and slow (soleus) twitch muscles in adult cats has been verified, and qualitatively similar differences foumd in adult rabbits and rats. In addition the conduction velocities of a motor nerve fibres to flexor digitorum longus, gastrocnemius medialis and tibialis anticus muscles have been measured in adult cats. Similar measurements (F.H.L. and soleus only), have been made in a series of kittens ranging in age from 8 hr. to 56 days. Although the conduction velocities in the newborn kitten are one-tenth those of the adult, the ratio of the conduction velocities of a motor nerve fibres to fast and slow muscles remains approximately constant from birth onwards. It is concluded that at birth the differentiation of the conduction velocities of the a motor nerve fibres to muscles destined to become fast and slow twitch muscles is complete, although at this time the muscles themselves are only beginning to differentiate in terms of contraction times. IT is usually considered that the hind limb muscles of the adult cat may be divided into two groups, fast twitch and slow twitch, on the basis of their isometric twitch and tetanus characteristics [Denny-Brown, 1929; Buller and Lewis, 1965 a]. These characteristics are at least in part dependent on the specific innervation of the muscle, so that when the nerves of a fast and of a slow twitch muscle are crossed surgically and allowed to regenerate, the twitch of the previously fast muscle is much slower, and that of the slow muscle much faster, than in unoperated specimens. The muscle contraction time is therefore determined at least in part by the type of nerve fibres innervating the muscle. In the newborn kitten the hind limb muscles do not show this clear division and, in muscles destined to become fast and slow contracting in the adult, the isometric twitch times-to-peak and the maximum rate of tension development in isometric tetani are similar in the two types. Differentiation then occurs as the animal grows older, values in the adult range being attained for these two quantities after about 6 weeks [Buller, Eccles and Eccles, 1960 a; Buller and Lewis, 1965 b]. In the adult cat, a motoneurones innervating fast and slow twitch muscles differ in certain characteristics, one of these being the conduction velocity of their axons [Eccles, Eccles and Lundberg, 1958]. It was shown by intracellular recording [Eccles, Shealy and Willis, 1963] that in kittens 3 weeks of age the adult pattern of motoneurone innervation was established, and that antidromic spikes were generally similar to those in adult motoneurones. However, * Present address: Department of Physiology, Yale School of Medicine. 333 Cedar Street, New Haven, Connecticut, U.S.A. 293

294 Ridge these workers did not separate lateral gastrocnemius and soleus nerves, and therefore no comparison was made between fast and slow innervations. It seemed of interest, therefore, to know whether or not the similarity in the timesto-peak of the isometric twitches of the muscles of the newborn kitten was associated with a similarity in the conduction velocities of their respective motor nerves, and this paper describes experiments made to determine this. The comparison was made between the motor nerve fibres to flexor hallucis longus (F.H.L. - as the cat has no hallux this muscle is strictly the medial head of flexor digitorum longus, but the accepted terminology is followed here) and soleus, as these two muscles have been most extensively used in comparing the mechanical properties of fast and slow muscles in the cat [e.g. Buller and Lewis, 1965 a]. In the adult rabbit [e.g. Vrbova, 1963] and rat [e.g. Close, 1964] also, fast and slow twitch muscles occur in the hind limb, and in this paper measurements of conduction velocities for motor nerve fibres to these muscles are also described. A preliminary note on some of these experiments has already been published [Ridge, 1965]. METHODS Animals and Anesthetics. - The experiments were performed on kittens in the age range 8 hr. to 56 days, and on adult cats, rabbits and rats. Kittens were anaesthetized with pentobarbitone sodium (Nembutal) using a dose level of 40 mg./kg. Due to the difficulty of anaesthetizing young kittens, the anaesthetic was diluted 1 in 5 with 0-9 per cent sodium chloride solution. Injections were made into the peritoneal cavity, and additional doses were administered similarly if necessary, dilution being 1 in 10. Adult cats, rabbits and rats were anaesthetized with Nembutal (40 mg./kg.), anaesthesia being induced first in rabbits with urethane (25 per cent) injected into a vein in the ear until the animal lost consciousness. Muscles and Nerves. - The muscles and their nerves were exposed, the spinal cord was opened, and dorsal and ventral roots of segments L5 to SI exposed. The roots were divided centrally and dorsal roots removed except in those experiments on adult cats where dorsal root potentials were recorded (e.g. in fig. 1). Pools for the spinal cord and the muscles and their nerves were formed with skin flaps, and filled with liquid paraffin, at 36-38 C. Prior to recording the compound action potentials from the spinal roots, nerves to be stimulated were divided from their muscles. Stimulating and Recording. - Stimulating and recording electrodes were of silver wire, with interelectrode distances of approximately 2 mm. (kittens and rats), or 3 mm. (cats and rabbits). The intensity of stimulation employed (of the order of 5V.) was approximately 1x3 x the minimum voltage required to give a maximal a motor fibre root volley, the stimulus duration being 30 or 100 usec. The cathode was always placed proximally. Stimuli were provided by an isolated pulse generator triggered from a Digitimer (Devices Sales Ltd.), and root volleys were amplified by a Tektronix 122 A.C. preamplifier (3 db points at 80 c./s. and 3 kc./s.) and displayed on a Tektronix 502 oscilloscope, the input being a.c. coupled (0 1,uF and 1 MQ). Temperature. - Rectal and pool temperatures were maintained in the range 36-38 C. by means of an electric blanket below, and radiant heat above, the animal. Measurement of Records. - Conduction Velocity. Records were projected onto graph paper by means of a photographic enlarger (approx. x 9), and measured. The time base on the records was derived from a 10 kc./s. crystal oscillator (Devices Digitimer). Conduction times from the stimulus artefact to the initial deflection of

Fast and Slow Muscle Innervation 295 the compound A.P. (as judged by eye) and to the peak voltage were measured. In young kittens especially, some of the action potentials tended to be diphasic (up to 20 per cent of total amplitude below the baseline), so that small errors (of at most 5 per cent) are associated with values derived from the peak latency. The results presented have not been corrected for these errors. Isometric Twitches. - In some kittens isometric twitch records were obtained. Tension was recorded by means of a Langham Thompson UF 1/4 unbonded strain gauge. Details of the recording and display system have been published [Buller and Lewis 1965 b]. RESULTS Adult cats. - In a short series of cats a comparative study was made of the conduction velocities of the a motor nerve fibres innervating soleus, gastrocnemius medialis, flexor hallucis longus (F.H.L.), flexor digitorum longus (F.D.L.) and tibialis anticus muscles. FHL SOLEUS 40] [0 VR... 401 [0 DR. *...... FIG. 1. Compound action potentials recorded in dorsal and ventral components of lumbar root 7, following stimulation of (a) F.H.L. and (b) soleus nerves (cat 7). VR, ventral root component; DR, dorsal root component. Each record consists of 5 superimposed sweeps. The three heights of time base marker are at 0 1, 0.5 and 1 m.sec. intervals respectively. Voltage calibration in jvs. Stimulus duration 0-. m.sec. (stimulus artefact at beginning of trace), and stimulus strength 1.3 xa maximum. Conduction distance 204 mm. for both nerves. Fig. 1 shows the compound action potential recorded from the dorsal and ventral roots when F.H.L. and soleus nerves were stimulated. This set of records was selected because the conduction distance was the same for the four records. It shows that the conduction velocity of the fastest fibres contributing to the compound action potential for soleus motor fibres is less than that for F.H.L. motor fibres; the values for the ventral root records are 106 and 81 m./sec. for F.H.L. and soleus respectively, giving a conduction velocity ratio VOL. LII, NO. 3.-1967 20

296- Ridge of 1-31. The conduction velocities for the ventral root action potential peaks in fig. 1 are 93 and 65 m./sec. for F.H.L. and soleus respectively, giving a conduction velocity ratio of 1-42. The difference found between F.H.L. and soleus motor nerve conduction velocities is not found in the sensory nerve fibres, and fig. 1 contains dorsal root action potentials to show this. Here the maximum conduction velocity of sensory fibres to F.H.L. is 112 m./sec., and to soleus 107 m./sec., giving a conduction velocity ratio of 1 05. For the dorsal root action potential peaks the conduction velocities are 97 and 93 m./sec. for F.H.L. and soleus respectively (ratio 1.04). In Table I the results for the complete series of adults are summarized. TABLE I. MEAN CONDUCTION VELOCITIES (M./SEC.) OF a MOTOR NERVE FIBRES IN THE NERVES TO HIND LIMB MUSCLES OF ADULT CATS, RABBITS AND RATS. VELOCI- TIES ARE CALCULATED FROM THE LATENCIES TO THE BEGINNING OF THE COM- POUND VENTRAL ROOT POTENTIAL (INFLEXION) AND TO THE PEAK VOLTAGE (PEAK). Conduction ±S.E. Conduction ±S.E. No. of Velocity for Velocity for Animal Animals Muscle inflexion peak Cat 7 F.H.L. 111.3±2.9 96-1±2 0 7 Soleus 89*7±2-0 73-8±2-3 6 GaStroC. med. 117.6±3-7 98 0±1±0 6 F.D.L. 114.2±2 0 95.8+1-8 6 Tib. ant. 104-8±3 5 90-0±2-9 Rabbit 3 F.H.L. and F.D.L. 86.5±5-7 75.3±5-3 3 Soleus 68-2±7-6 54-7±5.4 Rat 3 F.H.L. and F.D.L. 73.3±2-7 62.7±3*7 3 Soleus 64-3±4-5 52-2±3-3 When the conduction velocities of the a motor nerve fibres to F.H.L. were compared with those to soleus, it was found that there was a significant difference (p < 0 001) between them when calculated both from the latency of the beginning of the rising phase of the compound potential (fastest units) and from the latency to the peak. This comparison was made using the standard deviations of the differences in individual animals between F.H.L. and soleus conduction velocities. When the conduction velocity populations for the a motor nerve fibres of other muscles in Table I are compared with those of F.H.L. and soleus, it emerges that gastrocnemius and F.D.L. populations are much less significantly different from F.H.L. than from soleus. This is what would be expected, as both these muscles fall into the broad class of 'fast muscles'. However, the conduction velocity population of the fastest a motor nerve fibres to tibialis anticus falls between that of F.H.L. and soleus. Gordon and Phillips [1953] give times-to-peak of the isometric twitch of this muscle ranging from 19 to 24 msec., which would place it in the broad category of a 'fast muscle'. The slower contracting, quantitatively variable, anatomically separable part of this muscle that Gordon and Phillips also described would tend to prolong the timesto-peak of the whole muscle isometric twitch. Any slow conducting a motor nerve fibres in this portion would not be expected to affect the initial deflexion

Fast and Slow Muscle Innervation 297 of the ventral root action potential. The tibialis anticus results described here are similar to those of Boyd [1965], and would seem to contravene the general contention that fast-contracting muscles have fast-conducting a motor nerve fibres. However, as the quantitative relationship between individual motor unit times-to-peak and total tensions and the shape of the whole muscle twitch cannot be investigated except by full motor unit analysis, it would appear that only the relatively pure fast and slow muscles such as F.H.L. and soleus are useful for the type of gross analysis described in this paper. The conduction velocities of the fastest motor fibres to the muscles in Table I are compared with those to F.H.L. and soleus in Table II (p values). In two cats out of the 6 animals used for this comparative study, and in three further animals studied specifically, an inflexion on the falling phase of TABLE II. COMPARISON OF THE DIFFERENCES IN MAXIMUM CONDUCTION VELOCITIES OF THE a MOTOR FIBRES TO VARIOUS HINDLIMB MUSCLES, WITH THOSE OF FIBRES TO F.H.L. AND SOLEUS (P. VALUES). Comparing to F.H.L. to Soleus F.H.L. 0*001 Soleus 0 001 Gastroc. med. 0.01 0 001 F.D.L. 0.2 0-001 Tib. ant. 0.001 0 01 (h = 6 except for F.H.L. and soleus, where n = 7). the antidromic root action potential following stimulation of tibialis anticus nerve was seen. Boyd has observed a similar phenomenon (personal communication). Attempts to correlate the slower conducting fibres presumably responsible for this with the slow contracting portion of the muscle found by Gordon and Phillips, by attempting to stimulate selectively different populations of motor nerve fibres with varying intensity stimuli, were not successful. This observation was not pursued. Adult Rabbits and Rats. - In adult rabbits the maximum conduction velocities of the a motor nerve fibres to F.H.L. and soleus (2 animals) and F.D.L. and soleus (1 animal) were compared. Similar measurements were made in three adult rats, and the results are also presented in Table I. Comparing the conduction velocities for the fastest a motor fibres to fast (F.H.L. and F.D.L. values are combined) and soleus muscles, the significances of differences are: rabbit, p > 0001; rat, p > 005 (n = 3 for both). From this it is concluded that in both these species the difference in maximum conduction velocities for the a motor nerves to fast and soleus muscles occurs as in the cat. Kittens. - A series of fifteen kittens ranging in age from 8 hr. to 56 days was examined. In fig. 2 antidromic ventral root action potentials (2A, upper two records) in a 36 hr.-old kitten, resulting from stimulation of F.H.L. and soleus nerves, are shown together with the isometric twitches of the corresponding muscles

298 Ridge (2A, lower two records). The display of dots below the twitch records is described in the legend to the figure. It will be seen that the latencies of the root potential initial deflexions for F.H.L. and soleus are not the same (conduction FHL SOLEU S 50[........ ---------------- Yli it ii: 11111,11h U_ I111. t --AL A -~~~~i'r t v I100...I......I /J1 H_ "'iii! il/ "il B FIG. 2. Ventral root action potentials (above) and corresponding isometric muscle twitches (below) for F.H.L. and soleus in a 36 hr.- old (A) and a 56 day-old (B) kitten. The display of dots below the isometric twitch records represent quantitatively the following characteristics of the twitch. Reading from the left, the four groups of dots represent: First group.-initial tension of the muscle in g.; Second group (raised above base line).-time-to-peak tension in msec.; Third group (on base line).-time to half relaxation in msec.; Fourth group.-peak tension developed during the twitch in g. In A 1 dot=0.5 g. and 1 msec. In B 1 dot 5.0 g. and 1 msec. Voltage calibration in jxv. velocities 18-5 and 12-8 m./sec. respectively). However, in the isometric twitch records the times-to-peak of the two muscles are very similar (52 and 58 msec. for F.H.L. and soleus, respectively). In fig. 2B are shown the antidromic ventral root action potentials and

Fast and Slow Muscle Innervation 299 corresponding muscle isometric twitches of a 56 day-old kitten. In this case the maximum conduction velocities for F.H.L. and soleus motor nerves are 73-7 and 63-7 m./sec. respectively. By this age the muscle twitch times have adults 120 100 <. 75 0 50 25 x 0 0 3 1 5 20 50 AGE days A n 0a: 2 < 1-5 0 0 o 0 0 0 0 0 0 01 0-3 1 5 20 50 AGE days B FIG. 3. A. Conduction velocity of fastest fibres plotted against age (logarithmic scale) for kittens, soleus (crosses) and F.H.L. (filled circles) motor nerves. Mean values for seven adult cats shown to the left; bars represent the range. The line represents 0-86 m.fsec. increase in conduction velocity per day. B. The ratio of conduction velocities F.H.L. for the soleus kittens in A. Mean adult value and range to the left. Note that the F.H.L. motor nerve conduction velocity is always greater than that of soleus. become different, and the times-to-peak of F.H.L. and soleus are 21 and 60 msec. respectively. The results for the complete series of kittens are shown graphically in fig. 3. In fig. 3A are plotted the conduction velocities of the fastest motor nerve fibres

300 Ridge to F.H.L. (filled dots) and soleus (crosses) in m./sec., against the age in days of the kittens (on a logarithmic scale for convenience). The continuous line represents a linear relationship and corresponds to a rate of increase of velocity with age of 0-86 m./sec. per day. The points follow this line quite closely. Presumably at some age beyond 56 days the line would begin to flatten off towards the adult values, which are inserted to the left of 120 CD o100 X80 7 0 I1 60 0 0 340 20 A <80 B,60 40 C 20 g > 0 10 20 30 40 50 60 AGE days Fia. 4. A. Nerve length (F.H.L. only) plotted against age in kittens. Over this age range the relationship appears to be approximately linear, and the straight line was fitted by the method of least squares. B. Conduction velocity for fastest F.H.L. a motor fibres plotted against age in kittens. The line represents 0-8C m./sec. per day increase in conduction velocitv as in fig. 3, except that here the age scale is linear. the graph. It will be noted that for every animal the maximum conduction velocity for the motor nerve fibres to F.H.L. is greater than that for soleus, as was invariably the case in the adults. This point is made graphically in fig. 3B, where the F.H.L.: soleus maximum conduction velocity ratios are plotted against the age in days, with the adult values inserted on the left. It would therefore appear from these results that the conduction velocity difference between F.H.L. and soleus motor nerves seen in the adult is fully developed at birth, well before the full differentiation of the muscle speeds of contraction. This is so in spite of the fact that the absolute values of the maximum conduction velocities at birth are one-tenth of those in the adult. The increase in length of the nerve path also appears to be approximately linear in developing kittens, at least up to 56 days old. This is illustrated in fig. 4, where the length of nerve from ventral root to muscle for F.H.L. is

Fast and Slow Muscle Innervation 301 plotted against age in days. The line drawn in was fitted by the method of least squares. It must begin to flatten at an age sometime after 56 days, towards the adult length. The increase in conduction velocity with age for F.H.L. a nerve fibres is plotted on a linear scale in the same figure. However, the conduction times, in both F.H.L. and soleus motor nerves, while being prolonged in very young kittens, become similar to the adult values at an age of about 20 days. From that age on, conduction time is independent of nerve length in both kittens and adults. This is shown in fig. 5, where conduction times for F.H.L. and soleus antidromic root action potentials n 5 kittens adults 0~~~~ *@0 _. 3 ~~~~0 ^3 x x x x 0 ~~~00 1A x X X 0 0 50 100 150 200 250 Nerve Length FIG. 5. Conduction times plotted against nerve length in kittens and the six adult cats. Filled circles, F.H.L. and crosses, soleus nerve. are plotted against nerve length (maximum obtainable nerve paths were used for all latency measurements). This finding is related to the work of others in the Discussion. It implies that in kittens of 20 days or more old, and in adult cats, at least the peripheral motor conduction component of the total reflex time is fixed, and independent of the size of the animal. mm DIsCUssIoN It has become apparent during the last few years that the motor units making up a muscle are not all the same with regard to their contraction times, but rather fall within a certain range, the contraction time of the whole muscle being determined by the number, distribution, and relative tension developments of the various motor units. It is generally assumed, although as yet unproven, that the individual fibres making up any one motor unit in mammalian muscles have similar or identical contraction times one with another. This assumption is largely based on the smoothness of the isometric twitch records of single motor units (specific references below). It is also consistent with the dramatic changes in contraction times that occur when certain muscles have their nerves surgically crossed with the nerves of other muscles [Buller,

302 Ridge Eccles and Eccles, 1960 b]. This implies that, as all the fibres in a motor unit come under the same nervous influences, and as these are important in determining contraction times, the fibres are likely to be similar to one another in this respect. From the cross-innervation experiments it is apparent also that there are fundamental differences in the fibre mechanisms responsible for the different contraction times at least of the faster and slower motor units of cat hind limb muscles, and the possibility that there may be a complete range of motor unit contraction times does not detract from this knowledge. In the cat especially, and in the rabbit and rat and some other animals to a lesser degree, there occurs in the hind limb a great deal of segregation of motor unittypes into different muscles, so that it is valid to consider the broad classes of 'fast' and 'slow' twitch muscles in these preparations, bearing in mind that neither type is constituted of exactly similar motor units. As yet there are few full motor unit analyses of those hind limb muscles that have been used extensively as experimental material in the past, but McPhedran, Wuerker and Henneman [1965, soleus] and Wuerker, McPhedran and Henneman [1965, gastrocnemius medialis] have presented interesting analyses of soleus and gastrocnemius medialis in the cat. They showed that although there was considerable overlap between the two muscles, gastrocnemius medialis contained a number of units faster than any in soleus, and that soleus contained a number of units slower than any in gastrocnemius medialis. The idea of a continuous gradation of the slower motor units to the fastest, with a parallel distribution of other physiological and biochemical characteristics (such as lipid content and enzyme distribution) is almost certainly an over simplification. For instance, Stein and Padykula [1962, rat gastrocnemius and soleus] and other workers have classified muscle fibres into clear-cut classes on the basis of their straining reactions. However, none of these considerations alters the fact that the contraction time and other allied physiological properties (see Buller and Lewis, 1965 a and b) of 'fast' and 'slow' twitch whole muscles in the hind limb of certain animals, and particularly of the cat, become differentiated as the young animal grows up, and are dependent on their own specific innervations for their continued complete differentiation. It is therefore still valid to recognize these broad classes of muscle in these animals. In adult cats it has been shown that the conduction velocity of the fastest a motoneurones supplying fast and slow twitch muscles aredifferent, that to the slow muscle being slower than that to the fast [Eccles, Eccles and Lundberg, 1958; Boyd, 1965]. This relationship between conduction velocity of innervation and speed of isometric contraction, found for the two broad classes of fast twitch and slow twitch muscles in the cat hind limb, was also found to apply quantitatively at the motor unit level in a toe muscle of the cat. Bessou, Emonet-Denand and Laport [1963] showed that in the deep first lumbrical muscle, which has a small number of motor units (about 4-10), the times-topeak of the isometric twitches of individual motor units was linearly and inversely related to the conduction velocity of the motor unit axon. It is still not known whether this relationship applies to the motor unit populations of muscles such as soleus and medial gastrocnemius. Single motor units in these

Fast and Slow Muscle Innervation 303 two muscles in adult cats have been studied by McPhedran, Wuerker and Henneman (soleus) and Wuerker, McPhedran and Henneman (m. gastrocnemius) as referred to above, and the ranges of conduction velocities and contraction times are now known, but unfortunately in these workers' papers no data are presented for comparison of the conduction velocity of the a motor fibre and time-to-peak of the isometric twitch of the same xnotor unit. In the newborn kitten, where the times-to-peak of fast and slow muscle isometric twitches are very similar [Buller, Eccles and Eccles, 1960 a], one might expect the conduction velocities of the muscle a motor nerve fibres to be more nearly equal for muscles destined to become fast and slow in the adult. However, that this is not so is the main point shown by experiments described in this paper. The fact that the differentiation of the motor nerve fibres to F.H.L. and soleus in terms of conduction velocity is at least as marked in a kitten of a few hours old as in an adult would seem to imply earlier differentiation of the nerves (at least insofar as conduction velocity is a measure of this) than of the musculature (in terms of contraction times). It would be interesting to know whether the differentiation of the motoneurones in terms of synaptic connections found in the adult cat [Granit, Henatsch and Steg, 1956; Eccles, Eccles and Lundberg, 1958] is also found in kittens. Although reflex connections in kittens have been studied [e.g. Malcolm, 1953; Skoglund, 1960; R. M. Eccles, Shealy and Willis, 1963] there has been no comparison between fast and slow muscle motoneurones. Hursh [1939] reported a linear relationship between conduction velocity and fibre diameter in various nerves of adult cats and kittens (of unspecified ages) with a regression line slope of 6. He also reported that there was a linear relationship between internodal distance and fibre diameter in adult cats, and that data for a 16 day-old kitten fell on the same line. On the basis of this, conduction velocity may be considered as an indication of fibre diameter and internodal distance, at least down to an age of 16 days. However, as data for younger kittens are not available (nerve fibre diameters were not measured in this study), it would seem unwarranted to imply any structural basis for the conduction velocity data. Fig. 5 shows that the conduction time of the fastest a motor fibres in both F.H.L. and soleus nerves is greater in young kittens than in adults. This finding is in agreement with that of Skoglund [1960], who found that adult values of conduction time were reached at ages corresponding to a conduction velocity of the fastest afferents of 30 mn./sec. (gastrocnemius). This is similar to the results reported here, when conduction time reached the adult value at a nerve length of about 7 cm. (fig. 5), which corresponds to a conduction velocity of about 30 m./sec. for the fastest F.H.L. a motor fibres (fig. 4, a and b). In fig. 5, conduction time is plotted against nerve length in order to show the independence of conduction time and nerve length in older kittens (25 days and older) and adults. This is in spite of considerable differences in nerve length. The rate of increase of conduction velocity with age of the present series of experiments (about 0-86 m./sec. per day, fig. 3) agrees well with the results

301t Ridge of Skoglund, who found an increase of about 1 m./sec. per day for fastest gastrocnemius afferents. ACKNOWLEDGMENTS I should like to thank Professor A. J. Buller for his encouragement, Dr. P. Scott and her staff for kindly providing many of the kittens, Miss M. O'Vens for technical help and Mr. G. F. Stonard for his photography. REFERENCES BESSOU, P., EMONET-DE'NAND, F. and LAPORTE, Y. (1963). 'Relation entre la vitesse de conduction des fibres nerveuses motrices et le temps de contraction de leurs unites motrices', D.R. Acad. Sci. Paris 256, 5625-5627. BOYD, I. A. (1965) 'Differences in the diameter and conduction velocity of motor and fusimotor fibres in nerves to different muscles in the hind limb of the cat'. In Studies in Physiology, ed. D. R. Curtis and A. K. McIntyre. Springer-Verlag. BULLER, A. J., ECCLES, J. C. and ECCLES, R. M. (1960 a). 'Differentiation of fast and slow muscles in the cat hind limb', J. Physiol. 150, 339-416. BULLER, A. J., ECCLES, J. C. and ECCLES, R. M. (1960 b). 'Interactions between motoneurones and muscles in respect of the characteristic speeds of their responses', J. Physiol. 150, 417-439. BULLER, A. J. and LEWIS, D. M. (1965 a). 'The rate of tension development in isometric tetanic contractions of mammalian fast and slow skeletal muscle', J. Physiol. 176, 337-354. BULLER, A. J. and LEWIS, D. M. (1965 b). 'Further observations on the differentiation of skeletal muscles in the kitten hind limb', J. Physiol. 176, 355-370. CLOSE, R. (1964). 'Dynamic properties of fast and slow skeletal muscles of the rat during development', J. Physiol. 173, 74-95. DENNY-BROWN, D. (1929). 'The histological features of striped muscle in relation to its functional activity', Proc. Roy. Soc. B. 104, 371-411. ECCLEs, J. C., EcCLEs, R. M. and LUNDBERG, A. (1958). 'The action potentials of the alpha motoneurones supplying fast and slow muscles', J. Physiol. 142, 275-291. ECCLES, R. M., SHEALY, C. N. and WILLIs, E. D. (1963). 'Patterns of innervations of kitten motoneurones', J. Physiol. 165, 392-402. GORDON, G. and PHILLIPS, C. G. (1953). 'Slow and rapid components in a flexor muscle', Quart. J. exp. Physiol. 38, 35-45. GRANIT, R., HENATSCH, H. D. and STEG, G. (1965). 'Tonic and phasic ventral horn cells differentiated by post-tetanic potentiation in cat extensors', Acta physiol. scand. 37, 114-126. HURSH, J. B. (1939). 'Conduction velocity and diameter of nerve fibres', Amer. J. Physiol. 127, 131-139. MCPHEDRAN, A. M., WUERKER, R. B. and HENNEMAN, E. (1965). 'Properties of motor units; in a homogeneous red muscle (soleus) of the cat', J. Neurophysiol. 28, 71-84. MALcOLM, J. L. (1938). 'The development of reflex activity in the newborn kitten'. Abstr. XIX int. Physiol. Congr. 586. RIDGE, R. M. A. P. (1965). 'Conduction velocities of motor nerves supplying kitten hind limb muscles', J. Physiol. 176, 8P. SKOGLUND, S. (1960). 'The spinal transmission of proprioceptive reflexes and the postnatal development of conduction velocity in different hind limb nerves in the kitten', Acta. physiol. scand. 49, 318-329. STEIN, J. M. and PADYKULA, H. A. (1962). 'Histochemical classification of individual skeletal muscle fibres of the rat', Amer. J. Anat. 110, 103-124. VRBOVA, GERTA (1963). 'The effect of motoneurone activity on the speed of contraction of striated muscle', J. Physiol. 169, 513-526. WUERKER, R. B., MCPHEDRAN, A. M. and HENNEMAN, E. (1965). 'Properties of motor units in a heterogeneous pale muscle (m. gastrocnemius) of the cat', J. Neurophysiol. 28, 85-99.