## Abstract

The sensory innervation of cat hindlimb muscle spindles was studied by reconstruction, electron microscopy, and examination of teased, silver preparations to ascertain the form of the terminals and their distribution to bag<latex>$_1$</latex> (b<latex>$_1$</latex>), bag<latex>$_2$</latex> (b<latex>$_2$</latex>), and chain (c) muscle fibres. Reconstructions were made of two primary endings, one S<latex>$_1$</latex> secondary ending, and the branching of four primary and six secondary axons. For the silver analysis spindles were teased from 14 different hindlimb muscles, the largest samples being from tenuissimus, peroneus brevis, p. longus, p. tertius, superficial lumbrical, extensor digitorum longus, and soleus. Among 310 spindles examined, 40 lacked a b<latex>$_1$</latex> fibre. These were all portions of tandem spindles in which the b<latex>$_2$</latex> fibre was continuous from one capsule, where it was accompanied by b<latex>$_1$</latex> and c fibres, to another, in which it was accompanied by c fibres only. These have been designated 'b<latex>$_2$</latex>c spindle units' as distinct from 'b<latex>$_1$</latex>b<latex>$_2$</latex>c spindle units'. Counts of myonuclei in the primary regions of four b<latex>$_1$</latex>b<latex>$_2$</latex>c spindle units revealed 52-106 in the nuclear bags, b<latex>$_1$</latex> bags averaging 68, b<latex>$_2$</latex> bags 80. The average number in a myotube region was nine (range 6-12), and in a c fibre 24 (range 11-38). The reconstructions showed a close association between nucleation and innervation, but no constant relation between number of myonuclei and terminal contact area. They also showed that the largest cross-sectional areas of each bag fibre corresponded with the sites of S<latex>$_1$</latex> secondary ending innervation. Approximately 70-90% of the cross-sectional area of each bag fibre was occupied by myonuclei in the bag region, 30-50% in the myotube regions, and 10% in the region of secondary innervation. In b<latex>$_2$</latex>c spindle units the equatorial nucleation of the b<latex>$_2$</latex> fibre usually resembled that of a c fibre. The intramuscular diameter of Ia axons supplying b<latex>$_1$</latex>b<latex>$_2$</latex>c spindle units (mean 7.5 <latex>$\mu m$</latex>, range 3.4-12.8 <latex>$\mu m$</latex>, n = 213) was generally thicker than that of Ia axons supplying b<latex>$_2$</latex>c spindle units (mean 5.1 <latex>$\mu m$</latex>, range 2.2-9.0 <latex>$\mu m$</latex>, n = 37). The distribution of terminals by the first-order branches (usually two) of Ia axons to b<latex>$_1$</latex>, b<latex>$_2$</latex>, and c fibres was exclusively from heminodes, and was either segregated (b<latex>$_1$</latex> fibres supplied separately from b<latex>$_2$</latex> and c fibres, thereby resulting in separation of dynamic and static inputs) or mixed. Mixing was restricted most frequently to the dynamic input, and usually resulted from b<latex>$_1$</latex> fibres sharing a supply of terminals with a few c fibres. Distribution of terminals was usually segregated in tenuissimus and mixed in superficial lumbrical muscles, but in most muscles neither type of distribution predominated. Among 270 b<latex>$_1$</latex>b<latex>$_2$</latex>c spindle units, 32 had more than one b<latex>$_1$</latex> fibre, and 12 had primary endings formed by two Ia axons. Primary terminal systems supplied to bag fibres consisted of a middle portion, in which the terminals were arranged mainly as regular transverse bands, and portions at each end, in which they were disposed irregularly. In the silver analysis, b<latex>$_1$</latex> terminal systems in 151 b<latex>$_1$</latex>b<latex>$_2$</latex>c primary endings were distinguished from those supplied to b<latex>$_2$</latex> fibres by having more of their total length occupied by irregular portions (on average 57%, as compared with 33% in b<latex>$_2$</latex> fibres), and more bands per unit length in the middle. In the two reconstructed primaries the b<latex>$_1$</latex> fibres received 33 and 37% of the total terminal contact area, the b<latex>$_2$</latex> fibres 25 and 24%, and the c fibres 5-12% individually, 42 and 39% collectively. Primary endings supplied to b<latex>$_2$</latex>c spindle units were mostly irregular in appearance. The polar position of 351 secondary endings was S<latex>$_1$</latex> 253, S<latex>$_2$</latex> 79, S<latex>$_3$</latex> 15, S<latex>$_4$</latex> 3, S<latex>$_5$</latex> 1; 67.8% were distributed to b<latex>$_1$</latex>b<latex>$_2$</latex>c fibres (mostly as S<latex>$_1$</latex> endings), 20.8% to b<latex>$_2$</latex>c fibres, 6.3% to c fibres, and 5.1% to b<latex>$_1$</latex>c fibres. The intramuscular diameter of II axons terminating as S<latex>$_1$</latex> endings was generally greater (mean 3.9 <latex>$\mu m$</latex>) than that of II axons terminating as S<latex>$_2$</latex>-S<latex>$_5$</latex> endings (mean 2.9 <latex>$\mu m$</latex>); 75% of b<latex>$_1$</latex>b<latex>$_2$</latex>c II axons had diameters that fell within the lower part of the b<latex>$_1$</latex>b<latex>$_2$</latex>c Ia diameter range. Most II axons had two first-order branches; the distribution of terminals by the first-order branches of b<latex>$_1$</latex>b<latex>$_2$</latex>c II axons was usually mixed. Terminals were derived either exclusively from heminodes (as in most S<latex>$_1$</latex> endings) or from both heminodes and penultimate nodes. The mean length of 313 secondary endings was 348 <latex>$\mu m$</latex> (range 138-716 <latex>$\mu m$</latex>) as compared with a mean length of 359 <latex>$\mu m$</latex> (range 242-608 <latex>$\mu m$</latex>) for 151 primaries. In 83 b<latex>$_1$</latex>b<latex>$_2$</latex>c S<latex>$_1$</latex> endings the innervated portions of the bag fibres represented, on average, 42% (b<latex>$_1$</latex>) and 51% (b<latex>$_2$</latex>) of the total length of the ending. In 64% of the endings the b<latex>$_2$</latex> fibre received more terminals than the b<latex>$_1$</latex>. In the reconstructed S<latex>$_1$</latex> ending the b<latex>$_1$</latex> fibre received 8% of the total terminal contact area, the b<latex>$_2$</latex> fibre 17%, and the c fibres 16-22% individually, 75% collectively. Some muscles had fewer secondaries than others; the Ia:II ratio ranged from 1:1.2 (superficial lumbrical) to 1:1.8 (peroneus longus). The constant features of spindle sensory innervation that emerge from this study (e.g. the dense primary innervation of the b<latex>$_1$</latex> fibre) are discussed in the context of spindle development and in terms of their functional significance. The data on the branching of spindle afferents is related to the work of others on pacemakers and the manner in which nerve impulses are generated from the endings and propagated into the axons. We suggest that transduction may occur by a deformation of the sensory terminal owing to increased tension in the basal lamina. It is supposed that the permeability of the Na<latex>$^+$</latex> channels is affected by an intracellular messenger (probably Ca<latex>$^{2+}$</latex>) released from a bound state by increase in cytoskeletal tension. Reasons are given as to why the afferent innervating b<latex>$_2$</latex>c spindle units should be regarded as primary. The probable functional significance of these units is discussed, and some correlations are made between the function of certain muscles and the characteristics of their spindle populations.