For generations, Americans have been struggling to understand the theory behind the electoral college, all along not realizing there really wasn't one.

Rather, the strange system by which this nation chooses its president was a contrivance pieced together 200 years ago this week in Philadelphia by a group of tired and irritable men anxious to finish their work and go home.

What's more, its final acceptance by the Constitutional Convention was based in part on a disastrously flawed prediction about American politics: Many of the delegates appear to have voted for the system only because they believed that few presidential candidates would ever command a sufficient national following to win a majority of the electoral college votes. The choice of the president, they believed, would thus almost always fall to the Congress.

The projection was wrong: Only twice in American history (1801 and 1825) has Congress been called upon to pick the president.

The executive branch of government was unquestionably the least thought out of all, in part because the country had had no experience with such an office on a national level and had never held an election for any national officer.

The system most familiar to delegates was the one used by many of the states to select their governors -- appointment of the executive by the legislative branch of government. For the first half of the convention, this was also the favored device for choosing the president -- appointment by the Congress of the United States, as proposed by James Madison in his Virginia Plan.

As it turned out, that approach was acceptable to Madison and many of the large-state delegates only insofar as the legislative branch was constituted to their liking -- that is, so long as both chambers were based on proportional representation, giving the lion's share of power to the large states.

With the "Great Compromise" of mid-July, creating a House based on proportional representation and a Senate in which each state had an equal vote, these states lost control of the Congress. At this point, letting Congress elect the president suddenly seemed much less appealing to the large-state men, and soon Madison and his allies abandoned the plan.

With the support of small-staters, however, and without an alternative that garnered much support, the idea still was dominant in the convention as it began its 13th week in late August. On Aug. 24, however, the situation changed abruptly.

On that day, the convention took up the mechanics of a congressional election of the president. Two choices were available: The Congress could choose a president in the same fashion it was to enact legislation, on the concurrence of each house voting separately. Or, the two houses could meet as one and vote for a president as if they were a single institution, the victory going to the candidate favored by the majority of all the senators and representatives voting together.

Each approach had a profoundly different political impact. The first method -- each house voting independently -- meant that each house had to agree on a candidate for president. It preserved the Senate's ability to independently block a candidate just as it could block any individual piece of legislation. Because the states were to be represented equally in the Senate regardless of size, this approach favored the small states.

The second method -- a count of the preferences of all the members of the Senate and House -- allowed the large states to dominate the presidential election because of their dominance of the House, the larger chamber based on proportional representation. This approach, a joint ballot, was favored by the large states.

On Aug. 24, John Rutledge of South Carolina proposed adoption of the joint-ballot method, asserting that it would be "the most convenient mode of electing" the president.

The reaction sounded like a replay of the large-state, small-state conflict that had torn the convention apart in June and July. Roger Sherman of Connecticut was on his feet immediately, denouncing the joint ballot as a tool to deprive the small states of their due weight in the system.

Jonathan Dayton of New Jersey (then a small state) declared that "a joint ballot would in fact give the appointment" solely to the House of Representatives. The small states could "never agree" to it, he said.

The convention nevertheless approved the joint ballot as an amendment to the executive branch proposition, then being debated on the floor. That was the beginning of the end of the congressional election of the president, for on that day, the small states, too, quietly began to withdraw their support for the idea of a congressional election of the president.

As yet, however, no consensus had formed around an alternative. On Aug. 31, 1787, the convention did what it always did when in a quandary: it referred the subject of the presidency to a committee, made up of a delegate from each state, for private consideration.

The committee met in the upstairs library of the State House, and no formal notes of its proceedings have been discovered. But one of its members, John Dickinson of Delaware, left behind an account (still uncorroborated) of what he described as the climactic moment in the committee proceedings.

Dickinson had been ill at times during the convention and had missed the committee's early sessions. When he finally got to attend one of the later meetings, he discovered that the group had still not arrived at an alternative to election by the Congress, and was preparing to return to the full convention without one.

Dickinson had served as chief executive of both Pennsylvania and Delaware, and knew too well the consequences of a chief executive dependent for his survival on the good will of a legislative branch. Dickinson asked to be heard.

The proposed presidency, he said, had powers "so many and so great . . . that {I} do not think the people would be willing to deposit them with him unless they themselves would be more immediately concerned in his election . . . . The only true and safe principle on which these powers could be committed to an individual is that he should be in a strict sense of the expression the man of the people."

If the public rejected this part of the constitution, Dickinson warned, "the whole would be lost, and the states would have the work to go over again under vast disadvantages."

The committee needed no further prodding. "Come, gentlemen," said Gouverneur Morris of Pennsylvania. "Let us sit down again and converse further on this subject."

They sat down, and after some discussion, "James Madison took a pen and paper; and sketched out a mode for electing the president."

This was the most difficult challenge of the convention -- so many interests to be accommodated in a single office -- but Madison and the committee rose to the occasion.The committee stripped Congress of the direct power to choose the president, pleasing large-state delegates like Madison, who no longer had confidence in the legislative branch.

Also, the committee removed from Congress the power that the convention had previously given it to appoint Supreme Court justices and ambassadors, and transferred the appointment authority to the president, with the advice and consent of the Senate. Rather than choosing direct popular elections, the committee created the electoral college, to which each state would send a number of presidential electors equal to the number of seats it had in the House and Senate combined.

This was primarily for the benefit of the South, where a large proportion of the population, the slaves, was not permitted to vote. Thus, a popular election would have drastically curtailed the South's influence in electing the president. Slaves were counted (albeit at the rate of three-fifths person per slave) in proportioning seats in the House of Representatives and would be counted in the electoral college apportionment as well, giving the South substantially more influence. The committee then determined that when no presidential candidate achieved a majority of the electoral college vote, the Senate would choose the president from among the leading contenders. This was designed to make the small states happy, since in the Senate all states enjoyed equal status.

Two hundred years later, this concession may appear to be a meaningless sop. But in 1787, virtually all seemed to assume that rarely would a single candidate attain an electoral college majority.

(The choice will invariably "devolve" on the Senate, said Alexander Hamilton of New York. The Senate, said James Wilson of Pennsylvania, "will have in fact the appointment of the president." It "puts the appointment in fact into the hands of the Senate," said George Mason of Virginia, "as it will rarely happen that a majority of the whole votes will fall on any one candidate.")

The committee presented its proposal to the full convention on Sept. 4. The only real controversy in the debate that followed was over the wisdom of letting the Senate -- as opposed to the House -- select the president. To many, the Senate, with its advise and consent powers, already had too much authority. The proposal "would convert that body into a real and dangerous aristocracy," said Edmund Randolph of Virginia.

To remedy this, the convention decided that when the electoral college failed to achieve a majority for one presidential contender, the House would make the choice from among the leading electoral college vote-getters. Each state's delegation in the House, however, would have a single, equal vote -- regardless of the size of each state.

As Roger Sherman and others described it in the convention, the electoral college was to serve as a kind of nominating committee, proposing several candidates for the consideration of the House. It was on the basis of that understanding that he and the small-staters accepted it.

The formation of national political parties -- coordinating supporters of a single candidate in states throughout the country, making it possible for a contender to win a majority of the electoral votes -- began to undermine the old reality within a decade of the adoption of the Constitution. The growth of newspapers and then radio and television as political influences throughout the nation further contributed to the emergence of nationally known presidential contenders.

The electoral college itself -- though sometimes said to serve some higher national good -- was not so described in the convention. Rather it was pieced together helter-skelter to accommodate various interests.

Ironically, as Alexander Hamilton noted in the Federalist Papers, the mode of electing the president was "the only part of the system" of any consequence that had provoked no controversy during the ratification debates. "The process," he boasted, "affords a moral certainty that the office of president will seldom fall to the lot of any man who is not in an eminent degree endowed with the requisite qualifications."

NEXT MONDAY: The disaffected