Abstract

Cortical areas differ in the size and distribution of neuronal cell bodies, density, and distribution of myelinated axons, connections, and functional properties. We find that they also differ in the diameter of long corticofugal axons, with the thickest axons originating from primary motor, somatosensory, and visual areas and the thinnest ones from prefrontal and temporal areas. Since diameter is proportional to axonal conduction velocity, it can be inferred that action potentials issued from the different areas will be relayed to their targets at different speed. Conduction delays also depend on conduction distance. By computing conduction velocity and conduction distances, we found the longest conduction delays for the primary visual and temporal areas and the shortest for the premotor, primary motor, and somatosensory areas, compatible with the available electrophysiological data. These findings seem to establish a new principle in cortical organization relevant to the pathophysiology of neurological or psychiatric illnesses as well as to the speed of information processing in cortical circuits.

Introduction

Cortical areas are characterized by differences in the size and distribution of neuronal cell bodies, density, and distribution of myelinated axons, connections, and functional properties (Campbell 1905; Brodmann 1909; Felleman and Van Essen 1991). In particular, regional differences in the density and thickness of axons were noted by the early anatomists (Campbell 1905). Nevertheless, the implications and functional significance of areal differences in axonal architectonics remained obscure.

In this light microscopic study, we measured the diameters of axons originating in the primary and association areas of the macaque monkey and projecting through the telencephalic commissures, capsula interna, and to the thalamus or intracortically. We investigated how axon diameters relate to other morphological parameters of the areas of origin, notably, the degree of myelination, soma size of the parent neuron, and length of the axonal projection. For each area, we computed the conduction velocity expected from the diameters of the axons, hence their conduction delays, and from the length of the projections. The results show that each area is characterized by a different diameter/speed and length of axonal projections, which together generate different conduction delays.

Differences in the diameter of cortical axons might help to explain the selective involvement of fiber tracts originating in certain areas, notably the prefrontal and temporal areas, in neurological and psychiatric conditions. Differences in the conduction delays might indicate different speed of information processing in the various areas with consequences for cortical dynamics and hierarchy. On the whole, the results conform to the view that “neural information may … undergo significant transformations within the axonal component of the neuron” (Waxman 1975) and that transformation in the timing of neural processing is one of the computational operations performed by axons (Innocenti 1995, 2011).

Materials and Methods

Injections and Histology

Six adult male macaque monkeys (3 Macaca fascicularis and 3 Macaca mulatta; body weight: 4–7 kg) from the animal facility of the Department of Physiology and Pharmacology of the University of Rome SAPIENZA were used. Biotinylated dextran amines (BDA; Invitrogen, Carlsbad, CA) of different molecular weight were used to optimize anterograde versus retrograde transport (Reiner et al. 2000). Surgery was performed by a professional human neurosurgeon under strict sterile conditions. Animals were preanesthetized with ketamine (10 mg/kg, i.m.) and anesthetized with isoflurane (Abbott) through a constant flux of a mixture of isoflurane/air. Five animals received three to five 0.3–0.5 μL injections of BDA MW 10 000 (10% in 0.01 M phosphate buffer) at the cortical locations shown in Supplementary Figures 1 and 2 through a 5 μL Hamilton microsyringe (#85) with a sharp needle (P/N: 7803-05/00). After 17 days survival, the monkeys were deeply anaesthetized with ketamine (5–10 mg/kg, i.m.) and metomidine (30 μg/kg, i.m.) and perfused transcardially with isotonic saline followed by 4% paraformaldehyde in 0.1 M phosphate buffered saline (PBS). One animal (CCT6) received 3 times 0.2 μL injections of BDA, MW 3000 at each of the following locations: inferior bank of the sulcus principalis, crown of the precentral gyrus, crown of the postcentral gyrus, respectively, aimed at area 46, M1, and S1. After 9 days survival, the animal was sacrificed as above. Animal surgery, preoperative, and postoperative care were according to Italian (DL.vo 116/92) and European (Directive 86-609 EU) guidelines for animal experimentation on primates. The brains were postfixed overnight in the same solution, cryoprotected by immersion in 30% sucrose in PBS and sectioned frozen on the coronal plane, except for the corpus callosum (CC) of the injected hemisphere, which was cut sagittally in 5 of the monkeys. Section thickness was set at 60 μm, but it was later discovered that a malfunction in the microtome caused the sections thickness to be at 34 μm. Sections were reacted for BDA, and alternated sections were counterstained with cresyl violet or with the Gallyas method for myelin as in Caminiti et al. (2009).

Sampling and Measuring Diameters

All analyses were performed with the Neurolucida 7 software (MBF Biosciences, Williston, VT) and a digital camera--mounted Olympus BX51 microscope. The distribution of axons labeled with BDA was initially charted at ×260 magnification on sagittal sections of the CC. Axon diameters were subsequently measured at ×2900 on 112–600 μm wide probes, traversing the CC from dorsal to ventral in the regions of maximal density of labeled axons. In the measurement, the transversally cut axons profiles were approximated to circles whose size was incremented in 0.09 μm steps. To test the consistency of measurement, between 175 and 214 axons were measured in 3 nonconsecutive sections, at sites traversed by parietal axons in CCT1 and between 125 and 253 motor axons in CCT2. No significant differences were found in either animal. Since the measures in each animal were performed after an interval of more than 1 year, the results also prove the stability of the criteria used in the measurements. Measurements of tangentially cut axons entering the CC as well as of axons: 1) issuing from the injection site into the white matter, 2) entering the internal capsule, 3) entering thalamic nuclei ventralis lateralis (VL) and medialis dorsalis (MD), and 4) radiating from the injection sites into the gray matter were performed in coronally cut sections, in 2–3 nonconsecutive sections (Supplementary Fig. 6) at ×2900, as above. No correction for shrinkage was applied to the measurements. Since the same shrinkage should apply to the radial dimension of the axon and to its length, this would not affect the axon conduction measurements (below). The shrinkage of the BDA material was previously estimated to be 35% in the cat (Houzel et al. 1994). Retrogradely labeled somata of callosally projecting neurons were sampled in areas 4 and 46 in the animal (CCT6) injected with BDA MW 3000. The soma size was approximated by fitting it to a circle. The myelin density at the site of injection was estimated as ratios of transmitted light in layer 3 (inj) versus the underlying white matter (wm) each in 3 samples of 250 μm squares with the following formula: (inj/(inj + wm)) × 100.

Estimating Conduction Length, Speed, and Conduction Delays

Both the linear distance between the site of injection and CC and the modal curvilinear distance between the center of the injection site, the CC or anterior commissure (AC) midline, or the thalamus were measured (Tables 1 and 3). The first (not shown) was the distance between the section containing the center of injection and that containing the bulk of axons in the CC. The second was the length of the pathway reconstructed with the Neurolucida software from the center of injection (in layers 3–4) to the target by following the trajectory of the bulk of axons in serial sections. The curvilinear distance was adjusted to decompress the histological section back to the 34 μm values at cutting. Conduction velocity of the axon was estimated from the formula: Vc = (5.5/g)d (m/s) as in Caminiti et al. (2009), with g being the ratio between axoplasm d and fiber diameter D inclusive of the myelin sheath, with g set at 0.7. Conduction delay was estimated as: δt = L/Vc (μs).

Table 1

Diameter, conduction velocity, path length, and computed delay to the midline in all monkeys and injection sites

Transverse diameter (μm) Distance to 1/2 CC (μm) Velocity (m/s) Delays to 1/2 CC 
Origin Code Mean Median SD Mean Median SD Mean Median SD N 
Area 9 CCT2 0.68 0.62 0.22 13 442 5.38 4.87 1.74 2697 2759 681 634 
Area 9/46 CCT3 0.69 0.62 0.24 15 628 5.42 4.87 1.89 3140 3208 826 411 
Area F7 CCT2 0.86 0.72 0.36 12 579 6.72 5.65 2.84 2092 2224 605 426 
Area F4 CCT3 0.90 0.83 0.32 18 103 7.04 6.52 2.55 2870 2776 905 193 
Area 4 CCT2 1.04 0.90 0.48 16 901 8.19 7.07 3.81 2441 2390 914 253 
Area 4 CCT5 1.03 0.81 0.65 15 179 8.06 6.36 5.14 2552 2385 1247 1069 
Area 2 CCT3 1.13 1.00 0.46 21 151 8.90 8.10 3.60 2694 2614 909 166 
Area PEc CCT1 0.94 0.81 0.36 18 623 7.35 6.36 2.81 2870 2926 972 214 
Area PEa CCT5 0.80 0.63 0.41 19 649 6.28 4.95 3.23 3768 3969 1423 541 
Midtemp CCT4 0.75 0.64 0.34 29 398 5.85 5.03 2.64 5897 5846 2191 99 
Post-temp CCT4 0.77 0.64 0.45 23 229 6.05 5.03 3.54 4766 4619 1874 536 
Area MT/V4 CCT5 0.87 0.72 0.38 19 317 6.81 5.66 2.99 3304 3415 1188 592 
Area 17/18 CCT1 0.95 0.93 0.29 31 666 7.46 7.30 2.28 4578 4334 1180 273 
Ant temp AC CCT4 0.60 0.54 0.26 20 459 4.71 4.24 2.04 4828 4822 1355 268 
Transverse diameter (μm) Distance to 1/2 CC (μm) Velocity (m/s) Delays to 1/2 CC 
Origin Code Mean Median SD Mean Median SD Mean Median SD N 
Area 9 CCT2 0.68 0.62 0.22 13 442 5.38 4.87 1.74 2697 2759 681 634 
Area 9/46 CCT3 0.69 0.62 0.24 15 628 5.42 4.87 1.89 3140 3208 826 411 
Area F7 CCT2 0.86 0.72 0.36 12 579 6.72 5.65 2.84 2092 2224 605 426 
Area F4 CCT3 0.90 0.83 0.32 18 103 7.04 6.52 2.55 2870 2776 905 193 
Area 4 CCT2 1.04 0.90 0.48 16 901 8.19 7.07 3.81 2441 2390 914 253 
Area 4 CCT5 1.03 0.81 0.65 15 179 8.06 6.36 5.14 2552 2385 1247 1069 
Area 2 CCT3 1.13 1.00 0.46 21 151 8.90 8.10 3.60 2694 2614 909 166 
Area PEc CCT1 0.94 0.81 0.36 18 623 7.35 6.36 2.81 2870 2926 972 214 
Area PEa CCT5 0.80 0.63 0.41 19 649 6.28 4.95 3.23 3768 3969 1423 541 
Midtemp CCT4 0.75 0.64 0.34 29 398 5.85 5.03 2.64 5897 5846 2191 99 
Post-temp CCT4 0.77 0.64 0.45 23 229 6.05 5.03 3.54 4766 4619 1874 536 
Area MT/V4 CCT5 0.87 0.72 0.38 19 317 6.81 5.66 2.99 3304 3415 1188 592 
Area 17/18 CCT1 0.95 0.93 0.29 31 666 7.46 7.30 2.28 4578 4334 1180 273 
Ant temp AC CCT4 0.60 0.54 0.26 20 459 4.71 4.24 2.04 4828 4822 1355 268 

Note: SD, standard deviation.

Statistical Analysis

To evaluate statistical differences in axonal diameters and conduction delays, we first applied a nonparametric procedure (Kruskal–Wallis test, P < 0.05), since the distributions of samples generally deviated from normality. We then performed a multiple comparison of rank-ordered data to assess significant differences (Tukey–Kramer test, P < 0.05).

Results

Mapping the Cortical Surface into the CC

We performed mediolaterally spaced injections at corresponding anteroposterior levels in different monkeys (Supplementary Fig. 2). The injections had sharp boundaries, occupied volumes of 1–3 mm3 (Supplementary Fig. 1) as a rule restricted to the gray matter; each labeled between 99 and 1069 axons crossing the CC or the AC (Table 1).

Supplementary Figure 2 shows that the CC represents in its anterior to posterior dimension the anteroposterior location of cortical areas on the lateral surface of the hemisphere, confirming previous tracing experiments in the monkey (Pandya and Seltzer 1986) and in vivo tract tracing in monkeys and humans (Hofer et al. 2008). Instead, the dorsal to ventral dimension of the CC does not represent the medial to lateral location of the cortical areas. This occurs because cortical axons originating at a discrete cortical location course in a tight bundle toward the midline but defasciculate at the junction between the hemisphere and the CC (see Fig. 1 in Caminiti et al. 2009). Consequently, the axons cross within discrete elongated territories extending over the callosal thickness. The orientation of these territories follows the curvature of the CC: They are vertical in the body and become oblique or horizontal in the genu and in the splenium.

Figure 1.

Photomicrographs of callosal axons originating from prefrontal cortex (exp CCT2; A and C) and from area 4 (exp CCT5; B and D) in sections longitudinal (A and B) and transverse (C and D) to their trajectory. In (E), 2 axons from area 2 (exp CCT2) are shown at higher magnification and redrawn (blue) on the plane of the photomicrograph. Notice that axons from motor cortex are thicker than those from prefrontal cortex and that the axons undergo changes in diameter along their trajectory. The diameter was measured at regular intervals along the reconstructions (yellow circles). Scale bars in μm.

Figure 1.

Photomicrographs of callosal axons originating from prefrontal cortex (exp CCT2; A and C) and from area 4 (exp CCT5; B and D) in sections longitudinal (A and B) and transverse (C and D) to their trajectory. In (E), 2 axons from area 2 (exp CCT2) are shown at higher magnification and redrawn (blue) on the plane of the photomicrograph. Notice that axons from motor cortex are thicker than those from prefrontal cortex and that the axons undergo changes in diameter along their trajectory. The diameter was measured at regular intervals along the reconstructions (yellow circles). Scale bars in μm.

As shown in Supplementary Figure 2, axons originating at corresponding anteroposterior cortical levels but at separate mediolateral locations such as those from prefrontal or parietotemporal areas overlap. The axons labeled from the anterior temporal injection in ST2 course through the rostral part of the AC (Supplementary Fig. 2).

In the descriptions to follow the injection sites in areas 9 and at the 9/46 border will be usually referred to as “prefrontal,” those in areas 6 (F7 and F4) as “premotor,” those in PEc and PEa as “parietal,” those in V4 and MT (midtemporal) as “peristriate,” those in 17 ad 18, located near their common border, as “primary visual areas,” and those in ST2, ST3 and in PaAC/TPt as “anterior,” “middle,” and “posterior temporal” (locations according to Paxinos et al. 2000).

Diameters of Callosal Axons, Conduction Velocities, Path Lengths, and Interhemispheric Conduction Delays

The caliber of callosal axons clearly differs, according to their area of origin (Fig. 1). To quantify these differences, we measured the diameters of labeled axons both transversally, in sagittal sections of the CC, and longitudinally, that is, parallel to their trajectory at their entrance in the CC (Fig. 2 and Table 1). With both measurements, we found that the thickest axons originate in the motor, somatosensory, and primary visual areas and the thinnest axons in the prefrontal and anterior temporal areas with premotor, parietal, peristriate, and temporal areas falling in between (Table 1 and Fig. 2). Independent measurements by 2 of the authors and repeated measurements from the same author provided consistent estimates to at least the 97% level. Higher diameter values were found with the transverse than with the longitudinal measurements. The main reason for the difference is that the longitudinally coursing axons alternate thicker and thinner portions (Fig. 1), and the latter had been selected in the measurements. The longitudinal changes in axon diameter are not caused by the transport of the tracer and were seen also in myelin-stained axons. Axons labeled from different injection sites within each animal were statistically different (Kruskal–Wallis, P < 0.001). The same test applied across animals showed significantly thicker axons from motor, somatosensory, and visual areas and thinner axons from prefrontal and anterior temporal cortex.

Figure 2.

Top: means (with standard deviations) and medians (diamonds) of measurements of commissural axon diameters in transverse sections. Middle: curvilinear path lengths to the midline. Bottom: means (with standard deviations) and medians (diamonds) of computed conduction delays to the midline. In top and middle panel, origins of projections are ordered according to the anterior–posterior topography of the CC; in bottom panel, they are in the increasing order of mean delays. Path length was different for axons originating near the 17/18 border or in area 18 (middle panel), but the 2 axonal systems were indistinguishable at the midline (top panel), and the conduction delays were computed on the average of the 2 conduction distances (bottom panel).

Figure 2.

Top: means (with standard deviations) and medians (diamonds) of measurements of commissural axon diameters in transverse sections. Middle: curvilinear path lengths to the midline. Bottom: means (with standard deviations) and medians (diamonds) of computed conduction delays to the midline. In top and middle panel, origins of projections are ordered according to the anterior–posterior topography of the CC; in bottom panel, they are in the increasing order of mean delays. Path length was different for axons originating near the 17/18 border or in area 18 (middle panel), but the 2 axonal systems were indistinguishable at the midline (top panel), and the conduction delays were computed on the average of the 2 conduction distances (bottom panel).

The progressive increase in axon diameters between the prefrontal and the motor areas is due to the progressive addition of axons above 1.2 μm while the peak (mode) of the distributions remains the same or nearly the same (Fig. 3).

Figure 3.

Top: frequency of diameters in axons from prefrontal, motor, and one peristriate area. Bottom: frequency of delays to the callosal midline in axons from prefrontal, motor, primary visual, and posterior temporal areas.

Figure 3.

Top: frequency of diameters in axons from prefrontal, motor, and one peristriate area. Bottom: frequency of delays to the callosal midline in axons from prefrontal, motor, primary visual, and posterior temporal areas.

Conduction velocities were computed from the individual measurements of transverse axon diameters and faithfully reflect their distributions (Supplementary Fig. 3).

Conduction distances from the site of origin of the axons to the callosal midline were estimated from serial section reconstructions of the axonal pathways (Supplementary Fig. 4; Table 1). The longest distance (31.6 mm) was from areas 17 and 18, followed by the projection from the midtemporal injection (in ST3; 29.4 mm) and the shortest was from area F7 (12.6 mm). The conduction distances were shorter and fairly similar for sites of origin in the frontal lobe, including area 4. As a rule of thumb, conduction distances increased progressively from anterior to posterior areas, due to the relatively anterior location of the CC in the hemisphere (Fig. 2).

The conduction delays from the site of origin to the midline of the CC were computed from the conduction distance divided by the conduction velocity. This showed a progressive increase of conduction delays to the callosal midline between 2–2.4 ms for areas F7 and 4 to 4.5–5.9 ms for the temporal and primary visual areas (Fig. 2 and Table 1). Thus, the total interhemispheric delay can be estimated to vary roughly between 4 and 12 ms, although additional delays might be introduced in the terminal branches of the axons. The Tukey–Kramer test applied to rank-ordered data across experiments showed significantly (P < 0.05) longer delays for V1, and temporal axons, compared with all the other axons, and the shortest delays for premotor, motor, and somatosensory axons, with intermediate values for prefrontal and parietal axons.

Cortical areas differ not only for the diameter of the callosally projecting axons, for the conduction distances to the CC, and for interhemispheric delays that these parameters together generate but also for the range in the conduction delays (Fig. 3). The range of delays (distance between maxima and minima, outliers excluded) varied between 3.3 ms (in CCT2, area 9) and 9.7 ms (in CCT4 post-temp) under the effect of 2 competing factors: modal axon diameter and conduction distance. It decreased linearly with the first (Pearson's coefficient = −0.6) and increased with the second (Pearson's coefficient = 0.64). Data not shown.

The range of conduction delays within a cortical tract appears to have interesting consequences for cortical dynamics, and in the case of interhemispheric connections, it increases with brain volume from macaque to chimpanzee to human (Caminiti et al. 2009 and Discussion).

Correlation of Axon Diameters with Other Morphological Parameters

In other systems, such as the retinal ganglion cells (Stanford 1987), the terminal arbor of callosal axons (Innocenti et al. 1994), the sensilla of the crayfish antennule (Mellon and Christison-Lagay 2008), and the thalamocortical projections (Salami et al. 2003; for earlier data, see Waxman 1975), axon diameter and length appear to be adjusted to each other so as to maintain constant conduction delays. From what was described above, one can expect that this would not be the case for the interhemispheric connections. Indeed, we found no significant correlation between axon diameter and pathway length to the CC (Supplementary Fig. 5).

Since callosal connections are reciprocal, one might expect that cortical areas giving rise and also receiving thicker axons may be more myelinated. We therefore tested if a correlation exists between the degree of myelination of an area and the diameter of its callosally projecting axons. The density of myelin in layer 3 was measured as luminance of the transmitted light and normalized to the transmitted light in the underlying white matter. No significant correlation was found either (Supplementary Fig. 5).

In other systems, notably in the spinal motor neurons and in the retino-geniculo-cortical pathway, neurons with larger soma size give raise to thicker axons. To test if this relationship holds for cortical neurons, we measured the soma size of 339 neurons in area 46 and of 448 neurons in area 4, retrogradely labeled with BDA MW 3000 injected in corresponding contralateral sites (Fig. 4 and Supplementary Fig. 8A–C). All the labeled neurons were pyramidal cells, most of them in layer III with a few in layer VI. The mean diameter in area 46 was 10.6 μm and in area 4, 13.7 μm; the difference was highly significant (P < 0.0001; t-test). Therefore, the thinner callosal axons from the prefrontal areas originate from smaller neurons, and the thicker axons from motor cortex from larger neurons. Indeed, in a limited number of cases, the soma size of callosally projecting neurons and that of their axons, distal to the axon hillock at 50–70 μm from the soma, could be measured at high magnification. This showed a tendency to a significant (P < 0.001) linear increase of axon diameter with soma size (Pearson's coefficient = 0.73; Fig. 4).

Figure 4.

Distribution of soma diameters of callosally projecting neurons in motor and prefrontal areas. Inset shows the relations between somata and their axon diameters in individual callosally projecting neurons in both areas. The 2 are linearly correlated (Pearson = 0.73; P < 0.001).

Figure 4.

Distribution of soma diameters of callosally projecting neurons in motor and prefrontal areas. Inset shows the relations between somata and their axon diameters in individual callosally projecting neurons in both areas. The 2 are linearly correlated (Pearson = 0.73; P < 0.001).

Areal Differences in Axonal Diameters in Other Corticofugal Pathways

To test if differences in axonal diameters consistent with those found for callosal axons also exist for other corticofugal pathways, we measured 3 additional tracts (Supplementary Fig. 6). These were 1) the main axonal tract, that is, the bundle of axons which originates from the injection site and has entered the white matter, 2) the intracortical axons originating from the injection site and radiating within the gray matter (Supplementary Fig. 8D,E), and 3) the tract to the internal capsule. In each tract, the diameters of longitudinally sectioned axons were measured at 2–3 proximal to distal locations. This was done for injections in cortical areas for which more than one experimental case was available, that is, for injections in prefrontal, premotor, and motor cortex. Between 120 and 699 axons were measured in each tract.

This work led to 2 conclusions (Fig. 5 and Table 2). First, different tracts originating from each injection site consist of axons with different diameters. In general, the intracortical axons vary less and are thinner (range of medians across experiments and sites: 0.4–0.54 μm) than those of the main tract (range of medians: 0.36–0.89 μm) and of those directed to the CC (range of medians: 0.36–0.63 μm) or internal capsule (range of medians: 0.46–0.89 μm). Second, as for the CC axons, diameters increase from prefrontal to premotor to motor cortex for axons in the main tract and for those directed to the internal capsule while they are minimal and statistically nonsignificant for the intracortical tracts. The Tukey–Kramer test applied to rank-ordered diameter of axons of the main tract, and internal capsule within and across animals showed, as for callosal axons, significantly (P < 0.05) thinner axons from prefrontal than for premotor or motor cortex and thinner axons from premotor than for motor cortex, with the only exception in CCT2 where differences between axons of the internal capsule from premotor cortex were not significantly different from those originating from motor cortex. There were no differences for the intracortical axons except that the intracortical axons in CCT5 were slightly thicker than those from all the other injection sites.

Table 2

Longitudinally measured diameter of axons in 4 tracts origination from different areas in 3 different experiments (μm)

  Main tract Intracortical Callosal Int capsule 
Origin Code Mean Median SD N Mean Median SD N Mean Median SD N Mean Median SD N 
Area 9 CCT 2 0.43 0.36 0.19 699 0.49 0.46 0.21 908 0.46 0.43 0.13 423 0.51 0.50 0.19 191 
Area 9/46 CCT 3 0.49 0.46 0.15 182 0.48 0.46 0.12 149 0.43 0.36 0.16 226 0.46 0.46 0.13 481 
Area F7 CCT 2 0.57 0.50 0.22 322 0.51 0.40 0.20 726 0.54 0.50 0.24 154 0.61 0.50 0.24 120 
Area F4 CCT 3 0.66 0.55 0.33 419 0.49 0.46 0.22 279 0.54 0.46 0.30 153 0.64 0.55 0.30 233 
Area 4 CCT 2 0.72 0.66 0.34 203 0.54 0.44 0.31 569 0.63 0.51 0.31 262 0.68 0.59 0.33 188 
Area 4 CCT 5 1.02 0.89 0.48 315 0.66 0.54 0.28 193 0.87 0.63 0.49 255 1.01 0.89 0.51 233 
  Main tract Intracortical Callosal Int capsule 
Origin Code Mean Median SD N Mean Median SD N Mean Median SD N Mean Median SD N 
Area 9 CCT 2 0.43 0.36 0.19 699 0.49 0.46 0.21 908 0.46 0.43 0.13 423 0.51 0.50 0.19 191 
Area 9/46 CCT 3 0.49 0.46 0.15 182 0.48 0.46 0.12 149 0.43 0.36 0.16 226 0.46 0.46 0.13 481 
Area F7 CCT 2 0.57 0.50 0.22 322 0.51 0.40 0.20 726 0.54 0.50 0.24 154 0.61 0.50 0.24 120 
Area F4 CCT 3 0.66 0.55 0.33 419 0.49 0.46 0.22 279 0.54 0.46 0.30 153 0.64 0.55 0.30 233 
Area 4 CCT 2 0.72 0.66 0.34 203 0.54 0.44 0.31 569 0.63 0.51 0.31 262 0.68 0.59 0.33 188 
Area 4 CCT 5 1.02 0.89 0.48 315 0.66 0.54 0.28 193 0.87 0.63 0.49 255 1.01 0.89 0.51 233 

Note: SD, standard deviation.

Figure 5.

Means (columns, with standard deviations) and medians (diamonds) of longitudinal measurements of diameters for axons in 4 tracts, originating in prefrontal, premotor, and motor areas (data from three experiments).

Figure 5.

Means (columns, with standard deviations) and medians (diamonds) of longitudinal measurements of diameters for axons in 4 tracts, originating in prefrontal, premotor, and motor areas (data from three experiments).

Both the main tract and the internal capsule contain axons directed to a variety of targets, which are different for each area injected. The distance to the targets may compensate for the differences in axon diameters. Because all cortical areas send a feedback projection to their principal thalamic nuclei, we measured the distance from the prefrontal injection sites to the nucleus MD and from premotor (area F7 and F4) and motor cortex (area 4) to the nucleus VL complex. The sites of termination were identified by diffuse anterograde labeling, and they occasionally also contained retrogradely labeled cell bodies. However, these measures were complicated by the fact that the cortical projections defasciculate at some distance from the thalamus, and one must make sure that the axons are not intermixed with those originating from other injection sites in the same animal. Therefore, the measurements were performed separately and independently in 2 of the collaborating laboratories (by G.M.I. and S.T.) each measuring a different experiment. The prefrontal to MD distances found were 25.9 and 24.8 mm. The premotor to VL distances were 19.3 and 18.9 mm and the area 4 to VL 19.6 and 20.3 mm (Fig. 6 and Table 3). Clearly, the trajectory to the thalamus is longer for the projection from prefrontal than from motor cortex. This suggests that the conduction distance exaggerates the longer conduction delays to the thalamus which can be expected from the prefrontal cortex compared with the motor cortex, due to the thinner axons it projects to the internal capsule.

Table 3

Diameter, conduction velocity, path length, and computed delay to thalamic nuclei MD and VL

Longitudinal diameter juxta-TH (μm) Distance to TH (μm) Velocity (m/s)  Delays to TH (μs) 
Origin Code Mean Median SD Mean Median SD Mean Median SD N 
Area 9 CCT2 0.62 0.59 0.20 25 934 4.88 4.67 1.33 5694 5558 1475 238 
Area 9/46 CCT3 0.64 0.64 0.14 24 752 5.06 5.02 0.99 5071 4933 966 219 
Area F7 CCT2    19 257        
Area F4 CCT3    18 977        
Area 4 CCT2 0.71 0.67 0.27 19 615 5.55 5.05 1.85 3874 3883 1107 245 
Area 4 CCT5 0.83 0.74 0.44 20 306 6.53 5.94 2.93 3522 3416 1126 284 
Longitudinal diameter juxta-TH (μm) Distance to TH (μm) Velocity (m/s)  Delays to TH (μs) 
Origin Code Mean Median SD Mean Median SD Mean Median SD N 
Area 9 CCT2 0.62 0.59 0.20 25 934 4.88 4.67 1.33 5694 5558 1475 238 
Area 9/46 CCT3 0.64 0.64 0.14 24 752 5.06 5.02 0.99 5071 4933 966 219 
Area F7 CCT2    19 257        
Area F4 CCT3    18 977        
Area 4 CCT2 0.71 0.67 0.27 19 615 5.55 5.05 1.85 3874 3883 1107 245 
Area 4 CCT5 0.83 0.74 0.44 20 306 6.53 5.94 2.93 3522 3416 1126 284 

Note: SD, standard deviation.

Figure 6.

Top: frequency of diameters of intrathalamic axons from prefrontal cortex to the nucleus MD (experiment CCT3) and from motor cortex to nucleus VL (experiment CCT5). Bottom: lengths of pathways to thalamus from prefrontal, premotor, and motor cortex.

Figure 6.

Top: frequency of diameters of intrathalamic axons from prefrontal cortex to the nucleus MD (experiment CCT3) and from motor cortex to nucleus VL (experiment CCT5). Bottom: lengths of pathways to thalamus from prefrontal, premotor, and motor cortex.

However, the internal capsule contains not only axons directed to the thalamus but also those to basal ganglia, brain stem, and spinal cord. Therefore, at the thalamic level, a further axonal selection may occur. To test this possibility in 2 pairs of injections, in 3 experiments, we measured axons from prefrontal areas and from area 4 at their arrival in the thalamic nuclei MD and VL. The median diameter in MD was 0.64 and 0.59 μm, respectively, and the median in VL 0.74 and 0.67 μm (Fig. 6 and Table 3). As for the callosal axons, the increased diameter is due to the addition of axons above 1.2 μm while the peak (mode) of the distributions remains the same (Fig. 6). These values were similar but not identical to those found in the internal capsule projection. Although a selection of axonal diameters or axonal thinning might occur at the entrance to the thalamus, this appears not to be sufficient to compensate for the delays generated by the different lengths of the projections. The conduction delays to the thalamus were calculated in the same experiments assuming that the diameters along the whole pathway from cortex were the same as measured at the arrival in the thalamus. These were (medians) 5.5 ms to MD and 3.8 ms to VL in one couple of experiments and 4.9 ms to MD and 3.4 ms to VL in the other. The Tukey–Kramer test applied to rank-ordered data showed highly significant (P < 0.0001) differences between the different sites of injection. It should be noticed that the diameters were measured in longitudinally sectioned axons. Because transversally sectioned axons usually provide larger diameters, close to the electrophysiological data (above and Discussion), the conduction delays to the thalamus may have been underestimated. It might be relevant that the ratio between the delays to MD and, respectively, VL is almost the same (1.43–1.47) for the 2 couples of injections, although different animals were measured and independently by the 2 researchers.

Discussion

We found that different areas of the primate brain give raise to long corticofugal axons with different diameters. Instead the local myelinated axons, running within the gray matter had similar diameters in the areas explored and chosen because they exhibited maximal differences in the corticofugal axons. Although differences might exist among the unmyelinated axons which probably were missed in the present study, this might mean that the structure of the cortical modules “columns” (Mountcastle 1978) or “cortical output units” (Innocenti and Vercelli 2010) is invariant across areas because they perform similar computations while the long connections, those implementing Mountcastle's (1978) “distributed systems” are definitely area specific. The latter finding has potential consequences for brain pathology as well as for information processing in the brain.

Methodological Considerations

This study may have missed a few small unmyelinated axons, below the resolution of the light microscopy, that is, 0.2–0.3 μm. Indeed, previous electron microscopic (EM) work demonstrated the existence of small unmyelinated axons in the CC of the macaque (Swadlow et al. 1980). However, we chose the light microscopic approach for the following reasons. The number of small unmyelinated fibers in the CC of the macaque is below 10% and mostly close to 5%, with the exception of the genu/rostrum where it is close to 30% (LaMantia and Rakic 1990). Therefore, marginal underestimates, if any, of the thin axons, could be expected. Furthermore, these underestimates could not affect the main conclusions of the paper, which depend on the differential areal distribution of axons thicker than 1 μm, and on the length of axonal tracts, which together generate area-specific conduction delays in cortico-cortical and cortico-subcortical connections. Second, the results of our measurements turned out to be very close to those reported with EM. Indeed, for most of the CC, our final estimates differ by less than 0.1 μm from those of the EM study by LaMantia and Rakic (1990) in the same callosal sectors. A larger difference with axons 0.38 μm thicker in our study was found in sector 7 of LaMantia and Rakic (1990). It may be due to the fact that, in this sector, we sampled axons of somatosensory and motor origin, which now we know to be particularly thick, while other axons are certainly also coursing in the same sector. Finally, the conduction velocities and delays calculated from our light microscopic estimates are close to the available electrophysiological measurements, as elaborated below. Concerning the other projections, EM estimates of axon diameter appear not to exist, therefore, our analysis was restricted to the 2 ends of the spectra, the thinnest and the thickest axons, originating from prefrontal and motor cortex, respectively. It should be noticed that the light microscopic approach allowed minimizing the number of nonhuman primates used, an issue of concern in most countries.

A second limitation of the study is that it does not provide a comprehensive picture of the connectivity of all cortical areas nor of all projections from a single area. Work in progress might complete at least some of the missing information. At this stage, the study is meant to highlight what appears to be a new principle of cortical organization which further work from this or from other laboratories shall refine.

Areal Differences in the Diameter of Corticofugal Axons

The diameters of corticofugal axons do not correlate with the length of their trajectories, and this leads to differential conduction delays in the projections from the different areas, as discussed below. Unexpectedly, it does also not correlate with the degree of myelination of the supragranular layers of the areas of origin, which might rather be determined by the afferent and local connectivity. Instead, at least for the callosally projecting axons, their diameter correlates with the soma size of the parent neurons as suggested by previous work (Sloper and Powell 1979) and consistent with textbook knowledge in other systems, notably the magnocellular and parvocellular components of the retino-thalamic-cortical pathways and the spinal cord motor neurons. One would like to know if the diameter of cortical axons correlates with other morphological parameters, such as the spread of its terminal arbor, in particular the transmission compartment, the sector of the arbor which carries synaptic boutons (Tettoni et al. 1998). It would also be interesting to know if soma size of a cortical neuron relates to its threshold for activation.

The data thus far indicate that, in the monkey, association areas, notably prefrontal and temporal areas, send thinner corticofugal axons than primary motor, somatosensory, and visual areas. Abnormalities in axonal projections of association areas, in particular the prefrontal and temporal areas, are being reported in imaging studies of a number of major human neuropsychiatric pathologies, including autism, attention deficit hyperactivity disorder, schizophrenia, and Alzheimer (Innocenti et al. 2003; Di Paola et al. 2010; Kumar et al. 2010; Noriuchi et al. 2010; Whitford et al. 2010 and references therein). These abnormalities were often detected in the CC, whose topographical organization in humans resembles that of the macaque monkey (Hofer et al. 2008). Unfortunately, the imaging data cannot be ascribed to a precise structural alteration since they may be due to increased intercellular spaces, altered myelination or axonal caliber, etc. as discussed in Kumar et al. (2010). The fact that projections from prefrontal or temporal areas are more vulnerable to adverse conditions leading to neurological or psychiatric disorders might be due to their thinner axons, smaller cell bodies, or to some other properties associated with size. The vulnerability might also be related to the late and protracted maturation of those axons (LaMantia and Rakic 1990; Luders et al. 2010; Deoni et al. 2011). Currently, no conclusive data appear to exist on either association. However, abnormal maturation of corticocortical axons provides one of the possible explanations for the abnormal synchronous oscillations observed in schizophrenic brains (Uhlhaas and Singer 2011).

Areal Differences in Processing Speed

It seemed safe to predict that, since the axon diameter determines its conduction velocity, different cortical areas should relay action potentials to their targets at different speed. However, conduction delays also depend on the length of the projections, which vary across areas. When we computed these 2 parameters together, we found important differences in the conduction delays to the midline of the telencephalic commissures, in particular the CC, with the motor, premotor, and somatosensory areas being the fastest, in the order of 2–3 ms, the primary visual and temporal areas being the slowest, between 4.5 and 5.9 ms. The projections to other targets seem to follow similar trends. The conduction delays from the motor area 4 to the VL nucleus are faster than, for example, that from prefrontal cortex to the nucleus MD.

Considering the possible extrapolation of monkey data to human tractography, it may be useful to notice that a linear relation was found between conduction delays and length of the commissural and corticothalamic projections. However, positive and negative distances from the regression line can be expected, as in the monkey. They are the consequence of differences in axon diameters, causing projections from motor, somatosensory, and primary visual cortex to be faster and those from prefrontal and anterior temporal cortex to be slower than predicted by conduction distances (Supplementary Fig. 7).

The results on conduction velocity and delays are inferential because they are based on extrapolations from the morphology, although supported by well-established structural–functional correlations. They cannot readily be extrapolated to delays to the target neurons (transmission delays) since the diameters and conduction properties of the axonal terminal arbor in the gray matter and the synaptic delays are unknown. However, our conduction delays are close to those available from estimates in the primate and human literature (Fig. 7). For the visual cortex of the macaque, our data can be directly compared with the electrophysiological estimates of Swadlow et al. (1978) based on antidromic activation of neurons in the prelunate region, roughly corresponding to our V4/MT injection sites. Neurons were activated from the contralateral hemisphere (n 51) and from the midline of the CC (n 19). Their stimulation sites extended more ventrally in the prelunate region, while the CC electrodes in the splenium and isthmus of the CC activated axons whose origin, based on the present data, included the primary visual peristriate and temporal cortex. From the CC midline, they found delays of 1.2–5.9 ms (median = 3.2 ms) while we calculated 0.9–6.8 ms delays (median = 3.4 ms) for MT and V4 whose axons could not be separated at the midline. For the cortical stimulation, they reported interhemispheric delays close to the double of our estimates (7.0 vs. 6.8 ms). The conduction velocities and conduction distances calculated by Swadlow et al. (1978) are also very close to our own. Longer interhemispheric delays were reported for inferotemporal cortex neurons (27–30 ms; Ringo et al. 1994), activated through the AC, than what can be calculated from our data for the anterior temporal axons coursing through the same commissure (about 2 × 4.8 ms). It should be noticed that those values were obtained from post-stimulus histograms responses to natural stimuli in split-callosum and split-chiasm monkeys, which might have caused additional activation delays. The available data from the human literature appear to confirm the faster callosal transfer between the motor areas compared with the visual areas (Fig. 7). For the precentral cortex, the interhemispheric conduction time for inhibitory responses evoked by transcranial stimulation was between 10.4 and 12.4 ms (Cracco et al. 1989; Ferbert et al. 1992; Meyer et al. 1995; Boroojerdi et al. 1999), which is close to the 10.9 ms calculated from our human anatomical data (Caminiti et al. 2009). For the visual cortex, the interhemispheric transfer time measured as the difference between contralateral and ipsilateral responses is 16–20 ms (Lines et al. 1984; Terasaki and Okazaki 2002; Saron et al. 2003). The latency differences in interhemispheric conduction, between the motor and the visual areas, in humans are twice as large as in macaques. Our findings seem to fit the anticipation of Milner and collaborators that different portions of the CC transfer information at different speed (Lines et al. 1984).

Figure 7.

Comparison of conduction times to the midline calculated from anatomical data and estimated from electrophysiological experiments in macaques and humans. Further explanations in text.

Figure 7.

Comparison of conduction times to the midline calculated from anatomical data and estimated from electrophysiological experiments in macaques and humans. Further explanations in text.

The faster conduction in the motor and premotor projections is not restricted to the callosal connections, but it seems to apply also to the corticothalamic projection, at least when compared with the projection from prefrontal cortex. Interestingly, different conduction delays appear to exist within the thalamocortical projection, with faster conductions from the VA-VL nuclei than from the nucleus VM, in the cat (Steriade 1995). As discussed above, differences in conduction delays become particularly important in the large brains and might be attenuated or difficult to detect in the small brains due to the relative longer time spent by action potentials in their intracortical than in their white matter trajectories (Salami et al. 2003).

The data presented in this paper should not be interpreted as if conduction velocity channels are solely determined by the cortical site where they originate. It remains to be explored if, from a given cortical area, different conduction velocity pathways arise, directed at different cortical and/or subcortical regions.

The finding that different areas are interconnected with different conduction delays is puzzling in view of the fundamental role attributed to synchronous activation in coding distributed neuronal functions (for a recent review, see Uhlhaas et al. 2009). One might expect differences in conduction delays to limit the temporal precision, which can be attained in temporal synchronization. Whether this might be so, depends on a number of unknowns including the possibility that dynamic relays compensate for the temporal asynchrony caused by the conduction delays (Uhlhaas et al. 2009). On the other hand, conduction delays could provide a mechanism to establish robust phase differences in the processing dynamics of different cortical areas (Panzeri et al. 2010 and below).

It is probably of physiological importance that axon diameters and length of the pathways combine to generate a different spectrum of conduction delays within and across projections. Indeed, in simulated cortical networks, “delays give rise to a wealth of bifurcations and to a rich phase diagram, which includes oscillatory bumps, traveling waves, lurching waves, standing waves arising via a period-doubling bifurcation, aperiodic regimes, and regimes of multistability” (Roxin et al. 2005). Another interesting consequence of the spectrum of conduction delays is that it can operate as a filter limiting the oscillatory regimes of neuronal activity (Roberts and Robinson 2008; Caminiti et al. 2009), and it might also facilitate the transition to self-sustained patterns of neuronal activity (Bojak and Liley 2010). These considerations are particularly interesting in an evolutionary perspective since the enlarged brain volume was paralleled by a moderate increase in axonal size and only in a fraction of the axons, therefore expanding the range of conduction velocities (Olivares et al. 2001; Wang et al. 2008; Caminiti et al. 2009), while conduction distances increased more drastically. The 2 together, as we have shown, combine to maximize the range of conduction delays. Thus, a consequence of evolution may have been an increase in the range of oscillatory regimes and in the spatial-temporal patterns of cortical activity.

What is the meaning of the shorter conduction delays from motor, premotor, and somatosensory cortex than from the visual and most of the association cortex? Probably, all operations performed by the brain are carried out by processing performed in “distributed systems” (Mountcastle, 1978) involving multiple cortical areas in the 2 hemispheres, as well as corticofugal interactions with subcortical structures, in particular the thalamus. Therefore, from the conduction delays, one might infer a hierarchy of processing speeds in the brain, with motor and somatosensory cortex being the fastest. Motor and somatosensory areas might exchange more information per unit time than the visual or association areas. Furthermore, slower processing in visual or association areas might be continuously referenced to the faster somatomotor processing. It may be relevant that signals related to the onset of saccades appear to precede and modulate the timing of onset of neuronal responses to visual stimuli in area V1 of the monkey, although the effects may be mediated via the ponto-geniculo-cortical path rather than by corticocortical connections (Ito et al. 2011). More generally, one might speculate that operations performed in motor areas, such as the initiation of movement and the associated somatosensory feedback, might establish a primordial sensory motor self akin to Marr's (1982) “primal sketch” in sensory processing. Later operations will be referenced to this “sketch,” which they can bring to consciousness, validate, or modify, for example, by integrating visual experience as shown by manipulations of body ownership (reviewed in Tsakiris et al. 2007; Slater et al. 2009). It seems important that the callosal axons from motor and somatosensory areas are also the first to be myelinated in the monkey (LaMantia and Rakic 1990), suggesting that the primordial sensory motor self is established very early in life.

Supplementary Material

Supplementary material can be found at: http://www.cercor.oxfordjournals.org/

Funding

European Union contract # 029023, Paul BROCA II (G.M.I), and the Compagnia di San Paolo (S.T. and R.C.).

We thank S. Valentiniene for technical assistance, A. Vercelli for critical reading of the manuscript, and A. Battaglia-Mayer for help with the statistical analysis and the surgery. The authors are grateful to the Editors and technical staff of Cerebral Cortex who have allowed retraction of the first draft of this ms motivated by the discovery of a mechanical malfunction in the microtome, which required section thickness to be corrected from 60 to 34 μm. The gist of the paper remained unmodified, although the length of tracts with considerable anteroposterior trajectories, such as the callosal projection from areas 17 and 18 and the corticothalamic projections, was affected. Conflict of Interest: None declared.

References

Bojak
I
Liley
DT
Axonal velocity distributions in neural field equations
PLoS Comput Biol
 , 
2010
, vol. 
6
 pg. 
e1000653
 
Boroojerdi
B
Topper
R
Foltys
H
Meincke
U
Transcallosal inhibition and motor conduction studies in patients with schizophrenia using transcranial magnetic stimulation
Br J Psychiatry
 , 
1999
, vol. 
175
 (pg. 
375
-
379
)
Brodmann
K
Garey
LJ
Vergleichende Lokalisationslehere der Gosshirnrinde
Brodmann's ‘localization in the cerebral cortex’
 , 
1909
London
Imperial College Press
pg. 
300
 
Caminiti
R
Ghaziri
H
Galuske
R
Hof
PR
Innocenti
GM
Evolution amplified processing with temporally dispersed slow neuronal connectivity in primates
Proc Natl Acad Sci U S A
 , 
2009
, vol. 
106
 (pg. 
19551
-
19556
)
Campbell
AW
Histological studies on the localization of cerebral function
 , 
1905
New York
Cambridge University Press
pg. 
360
 
Cracco
RQ
Amassian
VE
Maccabee
PJ
Cracco
JB
Comparison of human transcallosal responses evoked by magnetic coil and electrical stimulation
Electroencephalogr Clin Neurophysiol
 , 
1989
, vol. 
74
 (pg. 
417
-
424
)
Deoni
SC
Mercure
E
Blasi
A
Gasston
D
Thomson
A
Johnson
M
Williams
SC
Murphy
DG
Mapping infant brain myelination with magnetic resonance imaging
J Neurosci
 , 
2011
, vol. 
31
 (pg. 
784
-
791
)
Di Paola
M
Spalletta
G
Caltagirone
C
In vivo structural neuroanatomy of corpus callosum in Alzheimer's disease and mild cognitive impairment using different MRI techniques: a review
J Alzheimers Dis
 , 
2010
, vol. 
20
 (pg. 
67
-
95
)
Felleman
DJ
Van Essen
DC
Distributed hierarchical processing in the primate cerebral cortex
Cereb Cortex
 , 
1991
, vol. 
1
 (pg. 
1
-
47
)
Ferbert
A
Priori
A
Rothwell
JC
Day
BL
Colebatch
JG
Marsden
CD
Interhemispheric inhibition of the human motor cortex
J Physiol
 , 
1992
, vol. 
453
 (pg. 
525
-
546
)
Hofer
S
Merboldt
KD
Tammer
R
Frahm
J
Rhesus monkey and human share a similar topography of the corpus callosum as revealed by diffusion tensor MRI in vivo
Cereb Cortex
 , 
2008
, vol. 
18
 (pg. 
1079
-
1084
)
Houzel
JC
Milleret
C
Innocenti
G
Morphology of callosal axons interconnecting areas 17 and 18 of the cat
Eur J Neurosci
 , 
1994
, vol. 
6
 (pg. 
898
-
917
)
Innocenti
GM
Exuberant development of connections and its possible permissive role in cortical evolution
Trends Neurosci
 , 
1995
, vol. 
18
 (pg. 
397
-
402
)
Innocenti
GM
Development and evolution: two determinants of cortical connectivity
Prog Brain Res
 , 
2011
, vol. 
189
 (pg. 
65
-
75
)
Innocenti
GM
Ansermet
F
Parnas
J
Schizophrenia, neurodevelopment and corpus callosum
Mol Psychiatry
 , 
2003
, vol. 
8
 (pg. 
261
-
274
)
Innocenti
GM
Lehmann
P
Houzel
JC
Computational structure of visual callosal axons
Eur J Neurosci
 , 
1994
, vol. 
6
 (pg. 
918
-
935
)
Innocenti
GM
Vercelli
A
Dendritic bundles, minicolumns, columns, and cortical output units
Front Neuroanat
 , 
2010
, vol. 
4
 pg. 
11
 
Ito
J
Maldonado
P
Singer
W
Grün
S
Saccade-related modulations of neuronal excitability support synchrony of visually elicited spikes
Cereb Cortex
 , 
2011
Kumar
A
Sundaram
SK
Sivaswamy
L
Behen
ME
Makki
MI
Ager
J
Janisse
J
Chugani
HT
Chugani
DC
Alterations in frontal lobe tracts and corpus callosum in young children with autism spectrum disorder
Cereb Cortex
 , 
2010
, vol. 
20
 (pg. 
2103
-
2113
)
LaMantia
AS
Rakic
P
Cytological and quantitative characteristics of four cerebral commissures in the rhesus monkey
J Comp Neurol
 , 
1990
, vol. 
291
 (pg. 
520
-
537
)
Lines
CR
Rugg
MD
Milner
AD
The effect of stimulus intensity on visual evoked potential estimates of interhemispheric transmission time
Exp Brain Res
 , 
1984
, vol. 
57
 (pg. 
89
-
98
)
Luders
E
Thompson
PM
Toga
AW
The development of the corpus callosum in the healthy human brain
J Neurosci
 , 
2010
, vol. 
30
 (pg. 
10985
-
10990
)
Marr
D
Vision
 , 
1982
New York
W.H. Freeman & Co
Mellon
D
Jr
Christison-Lagay
K
A mechanism for neuronal coincidence revealed in the crayfish antennule
Proc Natl Acad Sci U S A
 , 
2008
, vol. 
105
 (pg. 
14626
-
14631
)
Meyer
BU
Roricht
S
Grafin von Einsiedel
H
Kruggel
F
Weindl
A
Inhibitory and excitatory interhemispheric transfers between motor cortical areas in normal humans and patients with abnormalities of the corpus callosum
Brain
 , 
1995
, vol. 
118
 
Pt 2
(pg. 
429
-
440
)
Mountcastle
VB
Edelman
GM
Mountcastle
VB
An organizing principle for cerebral function: the unit module and the distributed system
The mindful brain: cortical organization and the group-selective theory of higher brain function
 , 
1978
Cambridge (MA)
MIT Press. p. 7--50
Noriuchi
M
Kikuchi
Y
Yoshiura
T
Kira
R
Shigeto
H
Hara
T
Tobimatsu
S
Kamio
Y
Altered white matter fractional anisotropy and social impairment in children with autism spectrum disorder
Brain Res
 , 
2010
, vol. 
1362
 (pg. 
141
-
149
)
Olivares
R
Montiel
J
Aboitiz
F
Species differences and similarities in the fine structure of the mammalian corpus callosum
Brain Behav Evol
 , 
2001
, vol. 
57
 (pg. 
98
-
105
)
Pandya
DN
Seltzer
B
Lepore
F
Jasper
HH
Ptito
M
The topography of commissural fibers
Two hemispheres—one brain. Functions of the corpus callosum
 , 
1986
New York
John Wiley & Sons Inc
(pg. 
47
-
73
)
Panzeri
S
Brunel
N
Logothetis
NK
Kayser
C
Sensory neural codes using multiplexed temporal scales
Trends Neurosci
 , 
2010
, vol. 
33
 (pg. 
111
-
120
)
Paxinos
G
Huang
X-F
Toga
AW
The rhesus monkey brain
 , 
2000
London
Academic Press
Reiner
A
Veenman
CL
Medina
L
Jiao
Y
Del Mar
N
Honig
MG
Pathway tracing using biotinylated dextran amines
J Neurosci Methods
 , 
2000
, vol. 
103
 (pg. 
23
-
37
)
Ringo
JL
Doty
RW
Demeter
S
Simard
PY
Time is of the essence: a conjecture that hemispheric specialization arises from interhemispheric conduction delay
Cereb Cortex
 , 
1994
, vol. 
4
 (pg. 
331
-
343
)
Roberts
JA
Robinson
PA
Modeling distributed axonal delays in mean-field brain dynamics
Phys Rev E Stat Nonlin Soft Matter Phys
 , 
2008
, vol. 
78
 pg. 
051901
 
Roxin
A
Brunel
N
Hansel
D
Role of delays in shaping spatiotemporal dynamics of neuronal activity in large networks
Phys Rev Lett
 , 
2005
, vol. 
94
 pg. 
238103
 
Salami
M
Itami
C
Tsumoto
T
Kimura
F
Change of conduction velocity by regional myelination yields constant latency irrespective of distance between thalamus and cortex
Proc Natl Acad Sci U S A
 , 
2003
, vol. 
100
 (pg. 
6174
-
6179
)
Saron
CD
Foxe
JJ
Simpson
GV
Vaughan
HGJ
Interhemispheric visuomotor activation: spatiotemporal electrophysiology related to reaction time
 , 
2003
Cambridge (MA)
MIT Press
Slater
M
Perez-Marcos
D
Ehrsson
HH
Sanchez-Vives
MV
Inducing illusory ownership of a virtual body
Front Neurosci
 , 
2009
, vol. 
3
 (pg. 
214
-
220
)
Sloper
JJ
Powell
TP
A study of the axon initial segment and proximal axon of neurons in the primate motor and somatic sensory cortices
Philos Trans R Soc Lond B Biol Sci
 , 
1979
, vol. 
285
 (pg. 
173
-
197
)
Stanford
LR
Conduction velocity variations minimize conduction time differences among retinal ganglion cell axons
Science
 , 
1987
, vol. 
238
 (pg. 
358
-
360
)
Steriade
M
Two channels in the cerebellothalamocortical system
J Comp Neurol
 , 
1995
, vol. 
354
 (pg. 
57
-
70
)
Swadlow
HA
Rosene
DL
Waxman
SG
Characteristics of interhemispheric impulse conduction between prelunate gyri of the rhesus monkey
Exp Brain Res
 , 
1978
, vol. 
33
 (pg. 
455
-
467
)
Swadlow
HA
Waxman
SG
Geschwind
N
Small-diameter nonmyelinated axons in the primate corpus callosum
Arch Neurol
 , 
1980
, vol. 
37
 (pg. 
114
-
115
)
Terasaki
O
Okazaki
M
Transcallosal conduction time measured by visual hemifield stimulation with face images
Neuroreport
 , 
2002
, vol. 
13
 (pg. 
97
-
99
)
Tettoni
L
Gheorghita-Baechler
F
Bressoud
R
Welker
E
Innocenti
GM
Constant and variable aspects of axonal phenotype in cerebral cortex
Cereb Cortex
 , 
1998
, vol. 
8
 (pg. 
543
-
552
)
Tsakiris
M
Schutz-Bosbach
S
Gallagher
S
On agency and body-ownership: phenomenological and neurocognitive reflections
Conscious Cogn
 , 
2007
, vol. 
16
 (pg. 
645
-
660
)
Uhlhaas
PJ
Pipa
G
Lima
B
Melloni
L
Neuenschwander
S
Nikolić
D
Singer
W
Neural synchrony in cortical networks: history, concept and current status
Front Integr Neurosci
 , 
2009
, vol. 
3
 (pg. 
1
-
19
)
Uhlhaas
PJ
Singer
W
The development of neural synchrony and large-scale cortical networks during adolescence: relevance for the pathophysiology of schizophrenia and neurodevelopmental hypothesis
Schizophr Bull
 , 
2011
, vol. 
37
 (pg. 
514
-
523
)
Wang
SS
Shultz
JR
Burish
MJ
Harrison
KH
Hof
PR
Towns
LC
Wagers
MW
Wyatt
KD
Functional trade-offs in white matter axonal scaling
J Neurosci
 , 
2008
, vol. 
28
 (pg. 
4047
-
4056
)
Waxman
SG
Integrative properties and design principles of axons
Int Rev Neurobiol
 , 
1975
, vol. 
18
 (pg. 
1
-
40
)
Whitford
TJ
Kubicki
M
Schneiderman
JS
O'Donnell
LJ
King
R
Alvarado
JL
Khan
U
Markant
D
Nestor
PG
Niznikiewicz
M
, et al.  . 
Corpus callosum abnormalities and their association with psychotic symptoms in patients with schizophrenia
Biol Psychiatry
 , 
2010
, vol. 
68
 (pg. 
70
-
77
)