(say in a deep, suave voice ala Sean Connery) My name is Bond, Covalent Bond. Here's the audio.
Historical Introduction: Events Before Lewis
Ionic bonds (called "polar bond" in those days) ruled the chemical world from their first suggestion in 1897 until the 1913-1916 time frame.
Remember, an ionic bond is formed by transferring one or more electrons from one atom to another.
Problem One: Electromeres
In 1901, it was realized that an electron could be transferred from atom A to atom B or from atom B to atom A. Therefore, the resulting compounds should not be the same. For example, NCl3 should exist in two forms, one with the nitrogen as +3 and one with the nitrogen as -3. These were called "electromeres" (after isomer) and the next 20 years were spent in fruitless searching for them.
Electromeres did not exist because, in many compounds, the electrons are "shared" in a covalent bond.
Also, it soon became clear the electrons were transfered in only one direction in an ionic bond. The positive ions were always positive, no matter what the negative ion was that was involved. (Needless to say, there are probably exceptions to this. We shall ignore them.)
Problem Two: J.J. in 1914
The first dissents to the polar bond orthodoxy began to appear in 1913 and in 1914, J.J. Thomson, the father of the ionic (or polar, to use the old style word) bond, himself became an apostate.
Thomson had spent many years studying the behavior of positive ions in gas discharge tubes. One such molecule was CO, carbon monoxide.
According to his theory, the bond in CO was ionic (to use the modern word). That meant the carbon was a positive ion and the oxygen was negative.
Hence, he reasoned, the positive ion beam he was studying should be made up of carbon. In fact, he found the beam to about evenly split between positive carbon and positive oxygen.
This was strong evidence against the idea of "intramolecular ionization" in molecules.
By 1913, Thomson had got around to publically accepting Rutherford's nuclear atom and in 1914, he published his new theory of chemical bonding. In it, he describes two types of bonds: the polar (our current ionic) and the nonpolar.
Thomson thought of the nonpolar bond this way: there was a tube of force (an attractive force) between the electron of atom A and the nucleus of atom B. However, the bond must be nonpolar, so only one electron cannot be involved. Thomson writes:
When the atoms are electrically neutral . . . for each tube of force which passes out of an atom, another must come in; and thus each atom containing n corpuscles [electrons] must be the origin of n tubes going to other atoms and the termination of n tubes coming from other atoms.
What Thomson was suggesting was two electrons were involved in a nonpolar bond. His idea is very close to the modern covalent bond and Thomson might today be known as the discoverer of the covalent bond, but for one mistake.
He proposed that the two electrons be considered as two bonds and so he proposed that the number of bonds in the structural formulas of chemists be doubled. That turned out to be wrong.