Since the visual system consists primarily of neurons, it is necessary to understand the neuron in order to provide a complete explanation of the visual system in animals. The literature of Neuroscience presents an unusual picture from a variety of perspectives. These will be highlighted briefly below.
Reviewing the most prominent and comprehensive research and pedagogical works in the field at the end of the Twentieth Century is enlightening. This literature is heavily weighted toward the morphological aspects of the neuron as opposed to the physiological (functional) aspects. This has been true for a very long time. There was an exploratory breakthrough in the 1950's that can not be denied. The work of Huxley, Hodgkin and Katz was pivotal. However, it does not appear to have been exploited by later work. Nearly 25 years later, in 1976, a major work appeared, "The Synapses" edited by Cottrell & Usherwood. It opened with a major chapter by Sir Katz that was mostly oriented toward the work up to the 1950's. The introductory chapter in a more recent major work, "The Axon" by Waxman, Kocsis & Stys (1995) is devoted to the work in the field up to the 1950's and was prepared by Sir Huxley. This is remarkable. Even 50 years later, the work of the above team is still considered one of the latest major breakthroughs in Neuroscience. Shepherd has played the role of dean of the neuroscience community for a long time. His Neurobiology has appeared in three editions up through 1994 but all based primarily on the original 1988 release. 1998 saw the publication of the text and atlas, Functional Neuroanatomy by Afifi & Bergman. Although profusely illustrated, it is based on the same dated material as the above works. It contains virtually no current information on signaling as an important function of the neurons. This work conflicts with many of the "Key Concepts" shown in the call-out boxes of that work. Few other fields of Science and no fields of Engineering exhibit the lack of growth of ideas at the fundamental level found in the Neurosciences.
Part of this slow growth of ideas at the fundamental level can be explained by the nature of electrolytic chemistry, the principle foundation of the functional aspects of the neuron. Examining the curricula of any major university shows that Electro-chemistry is a stepchild of Chemistry. Similarly, Electro-chemistry is a stepchild of Electrical Engineering. Electro-chemistry is a niche field commercially. The Electro-chemistry of the even more obscure area of the liquid-crystalline state of matter is even more obscure (although it has come into its own recently in high-tech consumer electronics). It is impossible to understand the functional aspects of the neuron and the nervous system without an appreciation of the field of electrolytic chemistry as it applies to liquid crystals.
Another major part of the phenomena can be explained by the fact that the laws of 20th Century Quantum Physics have yet to be embraced by the neuroscience community. The community generally speaks of Gaussian (Normal) statistics when most biological processes are controlled by Log-Normal statistics. When speaking of sub microscopic particle oriented events, it generally speaks in terms based on Boltzmann-Maxwell distribution law when the Fermi-Dirac distribution law is more appropriate when discussing the interaction of photons with matter. The "hole" of semiconductor physics is a concept that has found no natural home in the literature of the neurosciences.
The above difficulties appear to be related to the minimal mathematical training provided to biological students at least up through the 1980's. Without understanding the concepts of differential equations, it is virtually impossible to attack, or even comprehend, the dynamic aspects of the neuron. Without such a capability, the community generally falls back on simple pedagogical explanations of more complex concepts. The adoption of physical gates in the wall of a continuous cell membrane is such a crutch. The ion gate was originally a crutch to help mid-level biology students cope with the apparent transmission of ions through a membrane. The actual electrostenolytic processes when applied to semiconductor materials does not require the physical transport of ions through a membrane. The transport of electrons and holes provides the same chemical results.
It is still rare to find a major work in Neurosciences that recognizes the fundamentally analog nature of the neuron. The field is dominated by the view that most if not all neurons operate in the pulse mode, generating "Action Potentials" through a chemical process. This position seems to have evolved for two reasons. The larger and most accessible neurons are projection neurons found in the peripheral nervous system (PNS) and associated with motor activity. Motor neurons are overwhelmingly of the pulse signaling type.
However, from a functional perspective, an entirely different view emerges. As indicated in Waxman, et. al., the overwhelming number of neurons in any animal are actually associated with the central nervous system (CNS) and not the PNS. They give the ratio as 99.9% to 0.1%. The neurons of the CNS, which includes the retinas of the visual sensors, are fundamentally analog devices. Further more, many of the PNS nerves associated with sensory paths are in fact not of the pulse type. Only the projection neurons used to transmit signals beyond a few millimeters are of the pulse type. All other neurons involve analog signals represented by electrotonic potentials. There are some hybrid neurons that the above authors define as PNS-CNS compound axons.
A similar situation has evolved with respect to the interconnection of neurons. Because essentially all biologists are trained in chemistry and a bio-chemist is by definition a chemist, biological research has been built on a foundation of Chemistry, and certainly not on Electro-chemistry. As above, it is remarkable that the Neurosciences still rely on relatively simple chemical reactions to explain the operation of neurons, even in the face of the lament by Sir Katz in 1977: "...but I must confess that I am sometimes worried about whether, on our present evidence, we are justified in telling our students that chemical synapses are the predominant type in our central nervous system."
Although Huxley, Hodgkin and Katz were very careful in their original papers to circumscribe their work to the membrane wall of an axon, the community quickly generalized the work to relate to the total axon. And whereas, they showed the electrical parameters of the membrane wall to be virtually identical to those of the newly discovered (at that time) semiconductor diode, they and the neuroscience community proceeded to develop very simple electrical models to represent the membrane wall that are still in use today. There have been only weak attempts to explain the controlling mechanism for the various ionic currents that putatively flow through the membrane wall of an axon, and are represented by variable resistors in the equivalent circuits used to teach the subject matter.
To this day, the neuroscience community still employs passive electrical analogs to explain the operation of the neuron. Virtually no work has been reported that explains the operation of the neuron in terms of an active device. No work has been accepted that includes any active device within a neuron. Waxman, et. al. show an entirely passive network in their explanation for the operation of a mammalian myelinated axon, including the Node of Ranvier in 1995. This model is interesting only in that it is not a "minimal network" from a circuit synthesis perspective. In Circuit Theory, usually associated with electronics, a minimal network is required to elucidate the true characteristics of a network. The principle problem here is in the lack of definition of the variable resistors. How do they vary? What causes them to vary? Where is the governing input?
The presence of bio-energetic materials are known to be ubiquitous in the vicinity of neurons. However, all of the recent literature attempts to link these materials to the signal transmission function. Little or no material has appeared that would even suggest these materials play an electrostenolytic role in providing electrical power for an active electronic device within a neuron (or between neurons). This is unfortunate. When the concepts of electrostenolysis are applied to cytology, these bio-energetic materials are seen to take on an entirely different role.
The term PNS-CNS compound axon introduced above highlights another unusual feature of the Neurosciences. Shepherd has written a complete book on the "Foundations of the Neuron Doctrine" discussing the evolution of the concept of the neuron as an independent physiological unit. Apparently that battle rages on. Waxman, et. al. clearly consider the Axon as a fundamental physiological unit and Cottrell & Usherwood present the Synapse as a fundamental physiological entity. At some time, these different schools must come together with a satisfactory description of the signaling function provided by neurons that can be defended by the morphology (physiology). After all, neurons exist to provide signaling. They have no other reason to exist. Neuroscience needs an integrated view of the integrity of the fundamental physiological unit and the intimacy of the junction between such units.
This author believes that time has come. The answer lies in an active three terminal electrolytic semiconductor device found within and between all neurons associated with an individual signaling path. This device is known as an Activa. It is the electrolytic equivalent of the transistor. As Gutman et. al., have indicated in both 1981 and in 1983, a great deal of effort has gone into the search for such a transistor type of device in the non-biological portion of the organic chemistry community. However, nothing of the kind appeared. It has now appeared and been patented.
The literature of the neuron falls in two major categories, academic and clinical. The clinical literature depends on the academic for most of its discussion of fundamental mechanisms associated with neurology. The academic literature has developed by building on early concepts that are now out of date in many critical aspects. The academic literature of the neuron contains a great amount of experimental data and many discussions of how to interpret this data from a morphological perspective but no comprehensive model of the neuron at the fundamental functional level. Only recently, the view of the community has changed with regard to the projection neurons. Previously, the entire axon was viewed as a dissipative cable made up of distributed resistances and capacitances. More recently, it has been viewed as discontinuous, made up of lumped resistance and capacitances, with regeneration of the action potential at "hot spots." These "hot spots" are found near hillocks of ganglioon cells and at each Node of Ranvier--and result in a salutatory mode of signal transmission. Similarly, "hot spots" have been reported as occurring at various locations in the body of some neurons that are not associated with the nucleus. Recent references even report "hot spots" in the dendritic tree. This work will call on conventional (modern) electronic cable theory and demonstrate that the axon is primarily a non-dissipative cable consisting primarily of a coaxial cable exhibiting significant capacitance and inductance but negligible resistance. In this case, the various plasmas are the electrical conductors and the axolemma is the insulator.
A significant part of this problem has been the failure of the physiology community to provide a fundamental framework for the neural system that the morphologist can relate to. While maintaining an almost exclusively chemical perspective, the physiologist has not been able to account for the vast array of morphological features involving measurable electrical characteristics, specifically, electronically lucent and opaque regions.
There have also been extensive conceptual discussions attempting to relate the nervous system to a computer. This has frequently introduced the hypothesis that the overall neural system of an animal involves digital and/or binary signaling. Stacy & Santolucito have placed this discussion in context; "In summary, although the all-or-none nature of action potentials would suggest that information transfer in the nervous system is digital in nature, it actually is almost entirely, if not entirely analog in nature." However, this description remains superficial. It does not distinguish between binary and digital. It does not distinguish between projection and local signaling.
Kandel, writing in Kandel, Schwartz & Jessell repeats the conventional wisdom, based on morphology, that the neuron consists of four chief functional compartments–the cell body, dendrites, axons, and terminals. These would not normally be considered functional descriptors in other technologies. They are more rationally described as topographic, i. e. morphological, descriptors. He also continues the tradition of considering the cell wall of a neuron to be "excitable" even though our knowledge is continuing to detail the cell wall as a simple liquid crystalline film of triglyceride material without significant inclusions. Finally, He continues to explain the polarization of the cell in terms of a mechanical pump that transfers ions across the cell wall. These aspects of the conventional wisdom are not supported by this work. He also continues to describe the velocity of action potentials along an axon without differentiating between phase velocity and group velocity. His assertion that: "As far as is known, glia are not directly involved in information processing, . . ." is supported.
His statement that all nerve cells share the same basic architecture appears to be an oversimplification. He also claims that nerve cells differ most at the molecular level. This work will show that the photoreceptor cell exhibits a significantly different functional architecture and topology than most neurons, and that all neurons share the same detailed structures at the molecular level. It is also ironic that he supports the original assertion of Cajal (1900) that the feature that most distinguishes one neuron from another is shape. This feature may provide a simple, but misleading, classification of neurons but there are many others.
He also expresses an intent to focus on four basic features of the nervous system, including the mechanisms by which neurons produce signals and the patterns of connections between nerve cells. The discussion of the mechanism of signal generation is clearly superficial.
To avoid becoming embroiled in a very large area of physiology, this work will restrict itself to a certain class of neurons that are primarily found in the retina, with only a dalliance concerning other types of neurons.
Neurons, like everything else in physiology, can be viewed from several perspectives and be divided into separate categories based on each perspective. To bound this work, neurons will be divided into five major classes:
In general, a complete signal path passes through the sequence: #1, #2, #3, #4, (#3, #2, #3), #4, #5 where the bracketed group of neurons are located within the brain. This work will be principally involved in the first sequence of four types. These types are generally characterized as being 1-2 mm. or less in length.
One result of this work has been the elucidation of a possible morphological evolution of the neuron based primarily on electrophysiological requirements and the putative primacy of electrical signaling in the neural system. The figures in the text show how individual simple modifications to each form of neuron lead to a more sophisticated and functionally useful form. The proposed evolution is truly beautiful and verging on the miraculous from a system analysts perspective.
To further restrict the range of this work, it is useful to bound this work with respect to the types of synapses that will be discussed. Terminology is quite overlapping and therefore less than precise in this area. However, most investigators will accept the simple dichotomy in this area between the chemically transmitting synapse and the electrically transmitting synapse. Electrical rather than electrotonic is used here intentionally. Electrotonic implies a restriction to a continuous, i. e. non-pulse signal; a distinction which is not desired at this point.
It should also be pointed out that some investigators speak of signals which excite a neuron (implying that a signal that does not cause an action potential type of response is not important). In the case of vision, this is patently untrue. The interneuron as defined above does not generate an action potential; however, it supports the primary signal processing functions in the eye. Similarly, the interneuron does not accept action potentials as inputs.
Looking at the synapse from a different perspective, most investigators will accept the division of the synapse family into at least two groups based on the distance between the membranes of the two juxtaposed neurons. For reasons which will become obvious, this division will be made at a distance of 75-100 Angstrom in this work. This division is comfortably between the typical chemical synapse spacing of around 200 Angstrom and the commonly reported typical electrical synapse spacing of 20 Angstrom. The narrowness of the electrical synapse has led to the assignment of the less than specific name, "gap junction". Dowben makes a further distinction among the chemical synapses into those with a width of around 200 Angstrom and those between 500-1000 Angstrom which appear most frequently in the case of mylo-neurons.
To go further in the subdivision of classes of neurons is quite difficult and frequently misleading. Many attempts have been made to pursue the subject further on morphological grounds. However, as will be seen below, morphological categorization is fruitless unless the electrolytic topology of the neuron is brought into the discussion. Topology provides a clear indication of what are input structures and what are output structures--plus other types of structures--without depending on physical characteristics. The case of the amercine (Greek: no axon) cell is a good example; by simple topological arrangement, the input and output structures share parts of a common external cell wall, giving the appearance of a neuron with an input structure but without an output structure, even though the detailed morphological structures along this external wall look much like normal input and output terminal structures.
Based on the above discussion, this work will be limited to a discussion of electrical synapses utilizing a spacing of less than 75-100 Angstrom between the membranes of juxtaposed neurons. Furthermore, it will be limited to a discussion of sensory neurons, interneurons, hybrid neurons, projection neurons and their synapses. It will specifically exclude the discussion of mylo-neurons; and their synapses.
Because of its apparent electrical nature and importance, a new functional class of synapse will be discussed. This is the intra-cellular synapse, the functional part of the well known Node of Ranvier.
In the following chapters, it is shown that a neuron contains one or more sites which can be described as "active devices" in the terminology of the electrical engineer. These sites exhibit the phenomena known as "transistor action". They are therefore defined and described as "Activa", i. e. biological transistors. It will also be shown that transistor action can occur at the junction between any two membranes, whether external or internal to a given neuron, if the spacing requirement and other conditions are met. Thus, transistor action can occur between two neurons, and potentially more generally, between any two cells, i.e. a motor nerve cell and its associated muscle cells.
The specific location of these activa is defined for each type of neuron discussed here. The structure supporting this "transistor action" is also defined along with the resulting operating characteristics to the level required to understand the vision process. Much of the material needed to understand the fundamental operation of the biological transistor will be found in Appendix B.