Twenty-five years ago, while touring the border of his heavily wooded estate in Northwest Washington, the late Joseph E. Davies, then 81, discovered a gap created by three missing spikes in an iron fence along the northern edge of his property.

Davies, recollected one neighbor, seemed upset, but when he was reminded that the hole in his fence served as a "private gate" through which a number of nearby residents gained access to a path on the lower portion of the beautiful Tregaron estate, he relented. The unofficial "gate" remained open, and his Cleveland Park neighbors continued unobtrusively to enjoy the wooded pathways and open glades surrounding his big house on the hill.

The anecdote has the ring of a bygone era. Twenty-five years is a long, long time in the social history of a city. Elegant private estates such as Tregaron used to dot the hilly Northwest landscape. Today they are being replaced by clusters of row houses or, in some cases, high-rise dwellings.

Tregaron is no different. Davies, a former ambassador to Moscow and an intimate of two United States presidents, died in 1958, not long after his neighborly encounter at the fence. His complicated will left confused the future of the estate--named after a village in Wales whence his parents had emigrated in the 19th century--until it was divided and sold two years ago by order of a D.C. Superior Court judge.

The mansion, its outbuildings and immediate grounds now belong to the Washington International School. A development of 120 homes, snugly tucked into the sharp slopes of the land, is planned for the remaining 14.6 acres of the 20-acre plot stretching between Macomb Street and Klingle Road NW.

Anyone with an affection for faded grandeur will rue the change, but it is impossible to deny that great estates are no longer economically viable. The real question today is how the land can be used most benefically. In the absence of any practical public alternative--such as costly conversion to park land--the argument concerns not whether a given estate will be developed, but how.

Considered in this light, the Tregaron estate, despite the understandable disaffection of many of its longtime neighbors in Cleveland Park and Woodley Park, fares reasonably well. The developer's plan is far from perfect, but compared with the nature-destroying blitzkrieg of building now going on at Hillandale (the former Archbold estate on Reservoir Road) or the fenced precincts into which the old Glover estate (between Massachusetts and New Mexico avenues) has been divided, the Tregaron plan is almost heaven on earth.

This is to say that the Tregaron development plan is quite sensitive but it can be improved upon. It also underlines the general inconsistency and insensitivity of the crazy-quilt planning process in the District, split among various agencies of the District and federal governments and the semi-independent board of the Joint Committee on Landmarks of the National Capital.

The latter point is important not just for the vacant estates in the wealthy northwest section, but for the city as a whole. The lack of suitable and enforceable architectural, ecological and planning standards gives a scary unpredictability to development policies throughout the city. Citizens in Shaw, Anacostia or upper Northeast are every bit as prey to the process as the neighborhoods in upper Northwest, perhaps even more so because in Northwest, residents, on occasion, have been able to summon the wherewithal and the expertise to affect the results.

This clearly was the case with the Rockefeller estate near Foxhall Road, where the original plan, ruinous to the site, was significantly changed for the better by the analysis of Richard Ridley, an architect hired by neighbors, and the Arthur Cotton Moore designs for a "village of mansions" on the hillside that ultimately emerged from the process.

The Tregaron estate follows this general pattern, but it also is a very special, complex case. It is at once a historic property, containing a fine neo-Georgian mansion designed in 1912 by a notable architect (Charles Adams Platt of New York, who also designed the Freer Gallery of Art here); a beautiful hilly tract filled with magnificent trees; a border between Cleveland Park and Woodley Park; an appropriate entrance to the forested valley of Rock Creek Park, and an excellent location for in-city housing, close to Connecticut Avenue and subway connections.

As split by the court order, the estate is indeed an odd thing, a snaking residential zone superimposed upon a forest and a one-time golf course that surround a private school. As currently zoned it is odder still--a disaster waiting to happen. Current zoning regulations treat the site as if it were a flat suburban tract suitable for single-family houses, each with its own yard.

Citizens opposed to the development may arrogantly pretend otherwise, but to apply this zoning to the Tregaron estate would be to destroy it; one might as well build a steel mill on the site, or strip mine it. This is why the "planned unit development," a device to circumvent the awful effects of conventional zoning, is all-important to the site, and this is why the proposal of the Tregaron Corp. is, fundamentally, a healthy one.

In spirit the proposed development promises many excellent and necessary things: to save the existing mansion and contribute financially to its restoration; to provide an extensively engineered drainage system; to preserve more than 90 percent of the existing large trees, making them an airy green backdrop for architect Chloethiel Woodard Smith's alluring concept of terraced row houses, and to donate a large portion of the site, including trees, ponds and pathways, as a permanent easement to the National Park Service, while undertaking much-needed thinning and maintenance of the forest.

Recently, for the second time, the Joint Committee on Landmarks of the National Capital unanimously rejected the development plan, which had been scaled down to 120 units from an outrageous first proposal for 180 units. Given the adversary nature of the current planning system, this was perhaps a wise position. It gives the citizens, and good design, a better chance when the development proposal comes up before the Zoning Commission this fall, for unquestionably there is a need for further reduction in the number of units and the amount of paved area.

But the citizens' outcry and the joint committee's decision also can be interpreted as a preposterous exercise in snobbery and nostalgia. There is after all nothing intrinsically wrong with row houses and the location could not be better for a new community of in-city housing. What needs to be done now is to ensure that the best of the developer's promises will come true. If that job is done well, Tregaron in its new incarnation could become something of a useful model for other parts of the city.