On 2017 Aug 17, Alain Destexhe commented:

We are impressed by the time taken by Dr Barbour to make such comments (a search in PubMed shows a quite impressive number of comments by him on various papers). Less impressive is that Barbour comments contain basic physics errors. For example, Barbour's reasoning is made for electromagnetic waves, so yes it is true that radio waves will propagate near the speed of light in neural tissue. However, this confuses electromagnetic propagation with charge movement, which is at the basis of ionic currents in neurons. We suggest that Barbour follows a basic course in electromagnetism to convince himself of the fundamental difference betweem these two phenomena. This confusion between propagation of electromagnetic waves (photons) and the membrane currents (ions) leads to aberrant conclusions.

A basic course in electromagnetism will also teach that the 4th of Maxwell equations (Ampere-Maxwell law) contains a term about the density of the "displacement current" (dD/dt), and which precisely accounts for charge accumulation. This current is neglected in the traditional cable equations, which do not include Ampere-Maxell law. The traditional cable thus forbids charge accumulation in the medium as well as monopoles, by design, and thus cannot be used to make any reasoning about monopoles in neurons. This is why one needs to generalize cable equations, to include the displacement current and make them fully compatible with Maxwell equations, allowing charge accumation for example. Barbour seems not to have understood this, and describes the generalization of cable equations as "un-necessary".

Finally, a basic course of electromagnetism will also teach that electric monopoles are well known in physics, and that they can have various causes. The slow movement of charges is one of these causes, which predicts transient monopolar effects in materials. Our calculations show that the same may apply to neurons - there is a transient time at the onset of ionic currents, where there may be transient monopolar effects. We suggest this as a physical explanation for the monopoles that Riera et al. have observed in their experiments.

On 2017 Jul 31, Boris Barbour commented:

In this commentary about a paper by Reira et al

http://jn.physiology.org/content/108/4/956

Destexhe and Bédard make the naive suggestion that propagation of voltage in physiological saline would be slow enough to enable the creation of temporary current monopoles in violation of Kirchoff's law. They have also unnecessarily attempted to generalise equations governing current flow in brain tissue to allow for this possibility

https://journals.aps.org/pre/abstract/10.1103/PhysRevE.84.041909

The relevant quote from this commentary is:

"A first possible cause of monopolar contribution is that neurons may not strictly obey Kirchoff's laws. According to the standard model, the charges are assumed to move instantaneously (infinitely fast), which gives rise to the fact that the return current appears immediately. However, in reality, there is an inertia time to charge movement, because the mobility of ions in a homogeneous electrolyte is finite and is considerably slower compared to electrons in a metal."

The authors provide no usable quantification of this effect. In reality, the propagation of the voltage will be close to the speed of light and the associated delays will therefore be totally negligible on spatiotemporal scales of interest in biology. As succinctly explained on this wikipedia page

https://en.wikipedia.org/wiki/Velocity_factor

the propagation velocity will only be reduced from the speed of light by a factor of sqrt(er), where er is the relative permettivity. For pure water er = 80, and we can use this as an upper limit, because the addition of ions causes a modest decrease of er. The velocity of propagation would therefore be approximately

c/sqrt(80) ~ 3E8/sqrt(80) m/s = 3.4E7 m/s.

If we generously consider the brain to have a characteristic dimension of 1 m, even then propagation would require only 90 ns. Over a millimetre that would be 90 ps.

Thus, Kirchoff's law remains a spectacularly accurate approximation under physiological conditions and deviations from it cannot explain the apparent monopoles reported. In addition, there is no foreseeable benefit to neuroscience in modelling violations of the law.

On 2017 Jul 31, Boris Barbour commented:

In this commentary about a paper by Reira et al

http://jn.physiology.org/content/108/4/956

Destexhe and Bédard make the naive suggestion that propagation of voltage in physiological saline would be slow enough to enable the creation of temporary current monopoles in violation of Kirchoff's law. They have also unnecessarily attempted to generalise equations governing current flow in brain tissue to allow for this possibility

https://journals.aps.org/pre/abstract/10.1103/PhysRevE.84.041909

The relevant quote from this commentary is:

"A first possible cause of monopolar contribution is that neurons may not strictly obey Kirchoff's laws. According to the standard model, the charges are assumed to move instantaneously (infinitely fast), which gives rise to the fact that the return current appears immediately. However, in reality, there is an inertia time to charge movement, because the mobility of ions in a homogeneous electrolyte is finite and is considerably slower compared to electrons in a metal."

The authors provide no usable quantification of this effect. In reality, the propagation of the voltage will be close to the speed of light and the associated delays will therefore be totally negligible on spatiotemporal scales of interest in biology. As succinctly explained on this wikipedia page

https://en.wikipedia.org/wiki/Velocity_factor

the propagation velocity will only be reduced from the speed of light by a factor of sqrt(er), where er is the relative permettivity. For pure water er = 80, and we can use this as an upper limit, because the addition of ions causes a modest decrease of er. The velocity of propagation would therefore be approximately

c/sqrt(80) ~ 3E8/sqrt(80) m/s = 3.4E7 m/s.

If we generously consider the brain to have a characteristic dimension of 1 m, even then propagation would require only 90 ns. Over a millimetre that would be 90 ps.

Thus, Kirchoff's law remains a spectacularly accurate approximation under physiological conditions and deviations from it cannot explain the apparent monopoles reported. In addition, there is no foreseeable benefit to neuroscience in modelling violations of the law.