The Dow Theory says sell. If we are truly in a bear market for stocks, a sell signal may push investors into a panic period, the second stage of bear markets. Selling may beget selling. Prices accelerate their fall and the urgency to liquidate mounts.

The third and final phase of bear markets is typified by a spreading erosion of prices, as previous gains are erased from the secondary and tertiary issues as well as the cyclical and blue-chip giants which make up the Dow average of 30 industrial stocks.

The DOw Theory pulsated its last positive reading in July 1976, when the industrial and transportation averages made their last coincident highs at 1011.21 and 231.27, respectively. The industrials managed a further recovery high of 1014.79 on Sept. 21, 1976, and thereafter began the long, 13-month slide to just under 800 this week for a drop of some 20 percent.

The Dow transportation index had refused to confirm the Dow 30's bear trend until they broke under their previous reaction low of 203.85 on Monday. In part, the frequently strong transportation sector - bolstered by rotating investor liking for coal-hauling or asset-rich rails, profitable truckers or resurging airlines - were one of the several reasons why the persistent weakness in the Dow industrials has not been taken seriously enough by the majority of market strategists, professional money managers, or the general investing public.

Even now, the Dow utilities - which have never been part of the Dow theory - are ahead about 3 per cent on the year-to-date. The American Stock Exchange market value index is ahead some 8 per cent and the NASDAQ over-the-counter composite has shown almost a 3 per cent advance since January.

It's been a bad year for the big stocks, but now an uneasy feeling is spreading that the Dow Theory is telling us something.

How did we come to this? Charles Dow, the first editor of The Wall Street Journal, theorized in the 1890's that stock price movements have a relationship to future events. He saw the market's action manifested in three movements - primary, secondary and daily. Charles Dow saw in markets the equivalent of oceans in their ebbing and flowing characterized as bytides (the dominant long-term force), waves (secondary reactions) and ripples (daily fluctuations).

Dow's discovery of these rhythmic moves prompted The Wall Street Journal to construct and begin publishing in 1897 two sets of averages, the industrials and the rails. He concluded that these two sets of groupings were interdependent; manufacturing and commerce used transportation and both would prosper or subside in rough unison.

To reflect the evolution of transportation in America, the rail index was broadened to include other forms of transport on Dec. 22, 1969, to the chagrin of some Dow Theory purists. Yet the current Dow industrials and transports satisfy the original requirements.

The recognized forerunner of all historical and technical market studies, the Dow Theory frequently is criticized as being "too late," as it certainly is this year for holders of big stocks like those that carry the heaviest weight in the Standard & Poor's 500 index (down about 12 per cent this year), or the Dow 30 itself.

How this theory work? Any signal - up or down - must be confirmed by both averages to be authentic. It is sufficient if one average follows the other into new low(high) ground before the first average retracts its half of the signal. The first average to move retracts if it makes a new extreme move in the opposite direction before confirmation by the second average. Closing prices are always used.

What has its record been? Now that Charles Doe's successors - from early Wall Street Journal editor William Peter Hamilton and Dow theory student Robert Rhea to today's platoons of practitioners - would agree that the 1974-1977 bull market is over, should they be taken seriously? Where were they when the Dow was at 1000? Or even 900?

They would rejoin that Dow Theory confirmations by their nature - relying in closing penetrations of previous lows or highs by both averages -cannot catch bottoms or tops of amrkets.

But what the record shows in the years since 1897 is that the theory usually has kept the investor in or out of the market for the best part of most extensive moves. A study published recently in Barrons ("How Now, Dow Theory," Sept. 5, 1977) calculated that between 1920 and 1975, Dow Theory signals identified the direction of 68 per cent of the moves of the industrial and transportation averages and 67 per cent of those by the S&P 500 composite.

On the most mechanical basis - assuming positions were taken on one primary signal and not closed until the reversal was confirmed - the theory would have caught 35 per cent of the industrial and transportation moves and 29 per cent of the S&P 500's. So by pure performance, according to the strict rule, the Dow theorists have not done too well.

But those who exercise the judgmental factor and adopt varying market postures on directional inference presumably could have done much better.

Are we in a bear market? We have to await the next share price recovery before we can give the 1976-1977 experience its proper definition.

Remember that the Dow industrials have declined for nearly 10 straight months without a single rally of intermediate dimensions (5 per cent or more) - the first time in the 20th Century that we have experienced such an unrelieved decline in the market's most popular indicator.

In the meantime, an observer might note that if something acts like a bear, moves like a bear, looks like a bear, eats like a bear, smells like a bear and hangs around with other bears - maybe it is a bear.