Lawrence Phelps & Associates, A Corporation for Organbuilding

Effects of Wind Chest Design on the Speech of Organ Pipes


Since the completion of the new instrument, by Aeolian-Skinner, in The Mother Church, The First Church of Christ, Scientist, in Boston, many have remarked about its exceptionally well blended ensemble and the ease with which the whole instrument “speaks.” Many who have heard it, forearmed with the knowledge that special devices were used in the top boards on which some of the stops stand, have come to conclusions shared by the writer, who is the designer of the instrument, that these devices have made substantial contributions to this notable phenomenon. The writer is pleased to have this opportunity to describe these devices, and to present some of the facts and experiences which led him to design these special chests, and to further present some of his views on the above stated subject.

* * * * *

The voicing machine which the writer was privileged to use for several years in the Aeolian-Skinner plant was especially designed for the voicing of mixtures. It was built to hold three stops. The first top board was for principal pipes beginning with 4'C. The second top board was designed to hold a two rank mixture and the third was large enough to hold a three rank mixture. Early in his experience with this machine he began to notice that individual pipes voiced on the first top board seemed to have slightly different speech characteristics when placed on one of the other top boards.

This machine had regular pitman chest action and the first top board was normal, being like that shown in figure 63 on page 166 of the 5th edition of “The Contemporary American Organ” by Dr. William H. Barnes. The other two top boards were made in the usual way for mixtures. There was a horizontal boring in each, and the hole in the bottom was only deep enough to communicate with the horizontal boring, while at the top, in one case 2, and in the other 3 pipe holes were bored through into this horizontal boring. Then the horizontal boring was closed at the edges of the board, forming a key chamber. This meant that in the case of the two rank board, the wind, after entering this hole in the bottom, was communicated to two holes, rather than the usual single hole, via the horizontal key channel. The case was similar in the three rank board; while there was one chest valve and one hole in the bottom board, the wind was communicated by the key channel to three pipe holes in the top of the board.

In the voicing of mixtures, it was the writer's practice to voice all pipes on the front top board, that is the normal top board with pipe holes winded direct from the chest valves, and to place each rank into one of the other two top boards when completed, where they would remain to serve as samples until the entire mixture was completed. Of course, to play the ranks when they were placed on the multiple hole boards, it was necessary to block off the pipe holes on which no pipes were standing, in order not to rob the wind from the pipes which had been voiced. Although the speech of the pipes was carefully checked and regulated in the voicing process, it was noticed that when they were played on the chambered top boards they were considerably quicker in speech than they were on the normal board and that although at times this quickness was very slight, they were occasionally effected to the extent of being quite unsatisfactory in speech.

A friend played a large organ that had in it two well scaled mixtures. The voicing treatment of these stops was very inadequate and they were practically useless in the ensemble. The writer was asked if he would try to improve them. He accepted the challenge, completely re-voicing and very carefully regulating all of the pipes, but the effect of the stops was still not good. In this organ the wind chests had individual valves for every pipe, even including the mixture pipes, and no matter how many tricks of the trade were applied it was impossible to produce a well blended and cohesive sound from these mixtures.

After this experience the writer returned to the voicing machine determined to get to the bottom of this matter. On the next occasion for voicing a mixture, he set the pipes comprising low C of the mixture in the usual way on the first board. These pipes were then assembled on the mixture board at the rear of the machine and each pipe was adjusted to get the maximum well blended effect. Then these pipes were removed and put again on the first board and as suspected the blend was not as good. No matter what alterations were made in the strength and speech of the pipes, the blend could not be improved to equal the effect that was obtainable on the chambered mixture board.

These early simple experiences led the writer to two definite conclusions:

1. That expansion chambers, or chambers placed between the valve hole and the pipe foot, having larger volume than either the pipe hole or the valve hole, made it possible for the languids of pipes to be maintained in higher positions than was possible with pipes standing on direct valve holes, thus permitting the pipes to have easier speech without the rather disagreeable quickness that causes what he calls “gulping,” which is unavoidable when promptness is maintained on ordinary chests.

2 . That pipes speaking on common channels or key chambers, conveying wind to more than one pipe from a single chest valve, were more pliable and sound together in finer cohesion than pipes speaking on independent valve holes.

Later, it was found, that there was an optimum range of sizes for key chambers and that they must be considerably bigger than usual computations of the wind requirements of pipes would indicate. If the cross sectional area of the channel is not large enough the blending effect is hampered, and if it is too large the resulting clumsiness causes winding difficulties greatly outweighing the tonal advantages gained. The experience of revising and voicing the compound stops for the great Methuen organ with its colossal slide chests, and the later experience of caring for the instrument, together with observations of many other slide chests, were of considerable importance in further crystallizing the writer's ideas.

For the Great Organ of The Mother Church a Principal Chorus with 2/7th mouth pipes was decided upon because with these wide mouths greater amounts of tone could be produced than with narrower mouths. However, earlier experiences in the re-voicing of older 2/7th mouth pipes indicated that these pipes would not produce a really musical effect when placed on normal pitman chest top boards, as they then spoke with a gulping attack and were characterized by a hardness which was quite unpleasant. If the upper lip was raised to relieve this situation, the tone became far too broad and woolly. The designer felt that by abandoning Schulze type clubbed lips, using a much lower cut-up than the recommended by some authorities, and by placing these pipes on expansion chambers, the disadvantages which seemed to accompany these wide mouths in all previous American attempts could be completely overcome. Mr. Harrison agreed that these conclusions were logical, on a basis of his past experience, and before the contract was signed it was agreed that the Great chorus would have such pipes and that the 8' and 4' principals would be placed on expansion chambers, while the Scharf and Full Mixture would be placed on large key chambers. Sample pipes were made and placed on the slide chests in the Great of the old Hook and Hastings organ, which was still in use at the church, and this showed, beyond any shadow of doubt, that expansion chambers were really important to successful use of 2/7th mouth pipes.

It has always been the practice of Aeolian-Skinner to place mixtures on top boards similar to those used for them on the voicing machine described above. However, the size of boards in which it is possible to carry out this treatment, and also the size of the channels bored in them, is limited by the physical characteristics of wood. Large scale mixtures often require more than one top board. Thus such stops as the low pitched 4 rank Full Mixture on the Great, the Bombarde Grand Fourniture, and the six rank Swell Plein Jeu, would certainly have had to be placed on more than one top board, if this method of planting the pipes was used, in order to provide proper wind supply and enough space for the pipes to speak properly. Although this method of planting the pipes produces much better blending characteristics than is possible when each pipe has its own chest valve, the writer knew that much better blending characteristics would be possible if all of the pipes comprising any one mixture note could be placed over a common channel. Also he had often found that large scaled mixtures planted in the usual way did not get an ideal surplus of wind and he was determined that the large scaled pipes which he had specified would have an adequate surplus for the copious winding which he knew they would require. Therefore it was further agreed with Mr. Harrison that all ranks comprising a single note of any one mixture should stand on large key chambers.

The builders were now faced with the substantial problem of how to construct such devices into a standard pitman chest, in a manner that would stand the test of time, without putting their chest department completely off its production schedule. After much deliberation, in which literally every possible method of accomplishing this task was discussed, with careful consideration for the requirements of stability and uniformity, it was finally decided that the only practical method for building up the stock necessary for the proper construction of these channels would be to use wood in laminations.

It was accomplished as follows:

Laminated cores were built up out of slabs of wood approximately 1/4" thick. The length corresponded to that of the wind chests on which they were to be placed, 7' 6" for the Great, 7' 0" for the Positiv Cornet and 8' 6" for all others. They varied in width from about 6" for the Swell Cymbal, to 2' 6" for the Positiv and Bombarde Cornets. The finished thickness of the cores was 1 3/8" or 1 3/4" depending upon the estimated wind requirements for the stops. The necessary channels were cut parallel to each other across the width of these cores and to within about 1/2" of the edges. The channels were made as wide as possible and varied from about 1 1/2" in the bass to about 3/8" in the treble. After the channels had been smoothed and finished, the result was a “barred” core, or grid, similar to that used in slide chests. A layer of wood 1/2" thick was then glued to both top and bottom of these grids. The 1/2" boards glued to the top were bored in a manner corresponding to the pipe hole requirements of the normal top boards, while those glued to the bottom of the chest were bored to match the valve holes in the regular chest tables. In the very largest of these there were three holes in the bottom, corresponding to the three valves in the chest (which were used to insure adequate wind), and five holes in the top, corresponding to the five ranks of pipes. In every case the valve sizes were specified to be as large as the scale of the chest would permit. The enclosed grid or “sandwich” was glued to the normal table of the chest and great care was taken to be sure that the holes in the bottom of the grids matched very carefully the valve holes in the table of the chests. The pipes themselves were racked in the usual manner on regular top boards whose sizes were determined entirely in terms of convenience in handling. These top boards were screwed to the chest in the usual manner except that instead of being screwed to the table of the chest as shown in Barnes, op. cit. fig. 80, they were screwed to the grids which now formed the top surface of the chest.

* * * * *

In the shop all the C pipes of every stop were carefully set and these were used as samples by the voicers to guide them in the voicing of the remainder of the pipe work. The voicers were instructed to take great care that, even though the cut-ups of the pipes were much lower than normal, there should be no evidence of over quickness in the tone. The languids were to be kept high enough to avoid “gulping.” They were further instructed that the pipes of all stops which were to stand over expansion chambers or key chambers were to be voiced on the slow side. The voicers followed these instructions quite religiously and it was left to the tonal finishers to actually adjust the pipes to their permanent surroundings in the organ.

The pipes of the Swell were voiced first. These were placed in the organ and tuned and it was obvious before the tonal finishing had progressed very far that the cohesion of the mixtures was much better than usual. After these had been properly finished, regulated and finally tuned, there could no longer be any dispute as to whether or not key chambers produced a desirable effect upon pipes speaking in chorus. It was agreed by all that the easy singing effect produced by these mixtures had never before been heard from pipes standing on pitman chests.

The Great followed the Swell in the finishing schedule and the spectacular success of this flue ensemble immediately proved the value of key chambers to even the most skeptical observer.

The Hauptwerk was next on the schedule. In the Haupwerk no special attention had been given to the top boards except that the mixture boards were made thicker than usual and the channeled borings were made as large us possible. Other top board modifications were not planned because the scaling of this division was much smaller than the Great, the pressure was lower, and what were previously considered to be very successful results had been obtained elsewhere under these conditions.

The tonal finishing was accomplished from a temporary key board placed about in the center of the auditorium. The installation of the console and the tonal finishing of the Hauptwerk were completed at about the same time and it became possible to really hear the results of our efforts in the actual playing of organ music.

It soon became apparent that all was not well with the Hauptwerk. Mrs. Phelps complained that it seemed much more sluggish than the other divisions. To my own amazement I found that it was possible to play much more briskly on the Great Principal than it was on the Hauptwerk Prinzipal. This was the more surprising because by any standard of comparison a 43 scale “open” of the type of this Great Principal should certainly be considered lumbering when compared to a stop constructed in the manner of the Hauptwerk Prinzipal (45 scale) which in its voicing had been given every advantage towards encouraging spritely speech. The nicking was light, the cut-up was low and the pipes were made of tin, weighing a little on the light side. I found that nearly all of the Hauptwerk stops were really in this same sluggish condition. Even the tenor and bass octaves of the 2' Quinte and the 2' Superoctave were quite sluggish when compared with ranks of similar pitch in the Great.

It seemed certain to me that this difficulty was due to the fact that the Great pipes stood on modified top boards while those of the Hauptwerk did not, and that only top board modifications in the Hauptwerk could correct this difficulty. Quickening a few pipes by adjusting the languid produced a tone that was completely unsatisfactory, and it was evident that this solution to the problem was not going to be at all acceptable. Every detail of the chest action was carefully examined. Every known test was applied and it was deterimined that the difficulty was not action trouble of any sort.

All other possibilities having been exhausted, the writer's theory that the trouble could only be corrected by top board modifications was then pursued. A block of wood was made 3 1/2" square and 2" thick. A hole 2" in diameter was bored into the block to a depth of 12". On the opposite side of the block a normal 1/2" countersunk pipe hole was bored. This little block thus became in effect a portable expansion chamber. Tests were made on various Hauptwerk pipes by withdrawing the pipes from their holes, placing the little block over the pipe hole in the top board and then replacing the pipe so that its toe rested in the hole in the top of the block. Immediate improvement was noticed. Sometimes this improvement was slight, but in every case there was improvement and in most cases this was very marked. Having proved that expansion chambers would noticeably improve things, it was decided that these devices should be applied to 9 of the Hauptwerk stops. The top boards of these stops were sent back to the factory. As none of the top boards were more than 6 1/2" wide it was agreed that laminated stock would not be necessary. These top boards were not thick enough to allow for boring expansion chambers in their bottoms. Expansion chambers were therefore introduced in the following manner:

Planks were milled to the same dimensions as the top boards except that they were 2" thick. One plank was made to fit each top board. The centers of the top board pipe holes were marked on the planks and holes were bored completely through the planks on these centers. These borings, which were to be expansion chambers, were made as large as possible. They varied in diameter from about 3" to 1 1/2". The planks were screwed down to the chests in place of the top boards and the original top boards, unaltered, were then screwed in place on top of the planks. Thus expansion chambers had been created under every pipe hole. The pipes were replaced, and tuned.

The results were immediately remarkable. In fact several pipes actually had to have their languids raised, their speech had been so quickened by the application of the expansion chambers. No longer was there any feeling of sluggishness when playing or listening to the Hauptwerk and the ease of speech, which the designer was determined that this pipe work should have, was maintained. An interesting side point here is that the Hauptwerk 8' Bordun and the 16' Quintaden, as well as the 4' Kleingedackt, were considerably improved by the application of expansion chambers, a fact completely contradictory to theories held by some that note chambers do not affect the speech of covered stops.

Arguments as to the effect of chest design upon the tone and speech of organ pipes probably go back to the time the barred chest, the best form of which is of course the slide chest, was abandoned for a form of chest providing individual pipe valves. In 1906 Robert Hope Jones developed the standard Wurlitzer unit chest which used pallet valves and was equipped with very large expansion chambers. (The reader is encouraged to examine the excellent drawing of this chest in Barnes op. cit. fig. 73). So far as the writer knows, this was the last thought that was given to this matter before the decadence of the American organ set in. In its old form the slide chest could be constructed only by fine chest makers, well experienced in their craft, and it had to be built with great care. Because of this it was not possible to manufacture these chests fast enough to meet the demands of the growing industry. This was probably the real reason why the slide chest was abandoned. Many new forms of chests were devised which could be built more or less by production methods, where only the supervisory personnel need be skilled and the bulk of the work could be done by machine operators and young ladies with deft fingers. The situation was the same in Europe and all of the major builders there abandoned the slide chest in the construction of completely new organs. In this country many arguments were formulated against the slide chest, and indeed many poor examples of this chest were available. The growing fascination for gadgetry inspired by the success of electro-pneumatic mechanisms also contributed to the obsolescence of the slide chest.

After World War I the new German organ movement, headed by a group of organists and enthusiasts who recognized the decadent state of things, championed the return to first principles. Among other things they advocated the return to mechanical action and insisted upon the reestablishment of the slide chest as the only chest mechanism properly sympathetic to the type of voicing in which they believed.

In the early '30s, Walter Holtkamp, with one ear open to the arguments of the German organ movement and the other tuned to the fervent pleas of some excellent organists who were sincerely interested in effecting an American reform, built his first slide chest for a Positiv division which he added to the organ in the Art Museum in Cleveland. It stimulated a certain amount of interest, but unfortunately the bulk of opinion was that while chest design did effect the speech of pipes perceptibly, this effect was far too subtle to be of any importance and that what was really needed was scaling and voicing reforms. So Mr. Holtkamp has remained, until recently, the sole pioneer and champion of the cause of chest modifications as an avenue to better tonal results. The excellence of the slide chest built by Mr. Holtkamp certainly proves that many of the arguments against the slide chest are no longer valid.

The pitman chest is now predominant and is being used by even the country's smallest builders. It seems likely that the ventil chest may soon be abandoned, leaving the Austin “universal” wind chest as the only other style of chest construction practiced in this country. The new example in Boston is the first known American attempt to effect tonal improvement through modification of the pitman chest. The firm of Willis early recognized the significance of such modification in their practice of counter boring the pipe holes in the top boards of certain stops. This practice is still continued by the firm, but whether this is because of conviction or because of tradition seems to be a matter for question in the light of recent comments in these pages by Mr. Willis.

There is one observation that has been made as a result of these experiences in Boston which is most important. This instrument affords a rather spectacular demonstration of the fact that the effect of such chest modifications is not too subtle to be of fundamental significance and that the music loving public can readily recognize the difference. Many laymen and fine musicians who have always found it difficult to accept organ music due to the presence of the off-unison ranks have volunteered that they are not overly aware of the presence of such ranks in this instrument and that for the first time in their experience they actually enjoy listening to organ music. They found the old organs of the '20s completely uninteresting and most of the new organs a little too hard to take, but they find the present example “just right” (to use the exact words of several such commentators.)

* * * * *

It is the writer's view that his experience with the new organ in The Mother Church illustrates certain principles in the fields of chest design and pipe voicing that are applicable to small and , medium-sized instruments. They are summarized herewith.

With the price of organs constantly rising, a fact contributing substantially to the recent increase in the sale of electronic instruments, it is doubtful that anything contributing to further advances in the cost of pipe organs will meet with much popularity. Mr. Harrison has estimated that adequate expansion chambers can be applied to the pitman chest in normal production at an increase in cost of approximately 10 per cent per stop. While this would make for a substantial increase in the cost of small organs, the percentage of increase in organs of moderate size would not be so great but that it is reasonable to assume that those interested in better quality will not find this increase excessive. These devices should be applied to all of the chorus work, to broad strings, and to as much of the remaining pipe work as possible. Expansion chambers should definitely not be applied to reed pipes of the modern type, but they might be advantageously applied to reeds of the so-called baroque type provided these are voiced appropriately.

It is suggested that the manner of applying expansion chambers to the Hauptwerk, as described above, probably produces the best results with maximum of economy. Another method, which would perhaps produce fair results at slightly less expense, would be to use top boards at least 2" thick and to counterbore all pipe holes to a depth of at least 1 1/2", making the diameters of all counterborings as large as possible. This may be conveniently done to a pitman chest whose top boards are separate and completely independent of the action mechanism, such as those by Aeolian Skinner, and is possible with others by the addition of an extra set of special top boards.

The key chamber is by no means a cure-all. There is no substitute for voicing skill, finishing ability, and patience. However, when used with full understanding of their effect upon tone, with adequate time and proper personnel devoted to the voicing and finishing processes, these devices make possible certain excellent results which cannot be achieved when they are absent.

The practice of installing mixtures in such a way as to have every pipe of the mixture supplied by its own valve is deplorable and should no longer be tolerated. Attempts to justify this practice have been made on the ground that it makes tuning easier, so that less-experienced personnel can finish and tune the organ. However, it wastes the purchaser's money and space, penalizing him for the decision to use mixtures, rather than single-rank stops, while the results are not at all characteristic of good mixtures. This is one of the reasons why many mixtures which are “carbon copies” of successful examples with respect to pipe construction and voicing treatment produce mediocre results.

It is therefore extremely important that all the pipes comprising one note of a mixture speak upon a common channel and this principle must be applied not only to mixtures but to compound stops of all types such as sesqulalteras and cornets, et cetera. This can be successfully accomplished for small mixtures by using extra-thick stock for the top boards and boring liberal horizontal channels in these boards; but for large mixtures the method described above, using laminated cores, will be the most satisfactory.

* * * * *

In our attempts to get as far as possible from the stodginess of the organs of the '20s and the musical thought symbolized by these instruments, we have indulged in excessive tempi. In order that lightning-fast playing could be possible, and without substantial knowledge of the voicing practices and principles of the early European organ building masters to guide us, we have voiced our pipes extremely quick. This has resulted in a slight overblowing, especially of Principals, causing them to attempt to speak their octave at the attack. This produces a characteristically “gulpy” Principal Chorus with hard tone, and pipes with such individual stability that they cannot blend into a really cohesive combination when we draw two or more stops together. The smaller the organ the more objectionable this becomes. Either key or expansion chambers make it possible to voice Principal pipes so that their tone develops easily, like that of a violin as the bow is drawn across the string. Pipes of this sort have a warm brilliance without any tendency towards hardness, and an ability to combine with others of similar or different character in a way that is not otherwise possible.

It seems difficult, even for those well acquainted with the mysteries of flue-pipe voicing, to understand why pipes standing on expansion chambers of any sort require their languids to be higher than is necessary for pipes standing on a direct uniformly bored pipe hole supplied by a common disk chest valve. Because the chamber must be filled before the full wind pressure is available at the flue, it would at first seem logical that the languid should be lower in order to quicken the attack which must take place before the maximum pressure at the flue has been reached. However, when the languid is adjusted according to this theory it is found that when the full pressure does arrive the pipe is too quick and consequently overblown.

If there is no expansion chamber the wind reaches the flue in a quick puff or blast which causes considerable stiffness in the wind sheet at the moment of its first movement through the flue. This stiffness, together with the cross current toward the flue deflected across the bottom of the languid by the impact of the initial puff, causes a strong tendency for the wind sheet to start out from the body of the pipe in a direction away from the upper lip. Thus it is necessary that the languid exercise a strong influence on the direction of the wind sheet, and the edge of the languid must usually be considerably lower than the edge of the upper lip if the wind sheet is to be drawn into the upper lip far enough for the pipe to speak promptly under these conditions. The inertia of the air in the body of the pipe contributes substantially to this effect.

When an expansion chamber of adequate size is used there is no sudden puff at the flue and the wind sheet develops in a gradually accelerated movement until full velocity is reached There being no puff, there is no initial cross current under the languid. Under these conditions the first motion of the wind is straight up, directly toward the upper lip, and the inertia of the air in the body of the pipe is overcome by gentle persuasion rather than by sudden force. The pipe begins to speak easily and at once, with no necessity for forced speech. Great care must be taken not to over-nick the languid and indeed, if possible, to get along without nicks at all. This can be done properly only when the wind pressure is just right.

When the pipes are properly voiced the attack tone should consist primarily of the edge tone produced by the wind gently striking the upper lip. If the bevel of the languid is right (nearly vertical), and if the windway is properly adjusted the slight “chiffing,” found objectionable by Mr. Willis' “cultured ear” but undeniably present, in some form, in the attack tone of every wind instrument, accomplishes a definition for fast playing without the harmful hardness produced by the gulping of forced pipes. Until such pipes are consistently used in our small and moderate-sized instruments, Mr. Vente's remark about American organs that “the qualities and tones of their pipes are mostly romantic,” appearing in these pages in the Summer Issue, will remain true. These gently speaking, slightly percussive pipes work together to produce a greater variety in color, a more closely knit ensemble, better combination effects when stops of contrasting color are used together, and tones with greater warmth and intensity. This type of tone invariably meets with greater musical success and has a strong fascination for the man in the pew, who is often left completely cold by tone produced with pipes voiced according to other methods. I have seen considerable and conclusive evidence of this which no amount of argument to the contrary can refute. Regardless of any theories to the contrary the man in the pew is intelligent; his opinion does count and his enthusiasm should be a measure of the organ builder's success.

* * * * *

While it is true that pipes voiced in this manner and planted on chests equipped with individual expansion chambers for each pipe, with only the mixtures standing upon key chambers, will produce markedly better results than is possible without such chambers, even in the smallest instrument, the writer must state here his personal conviction, based upon his present experience, that there is no substitute for the slide chest if tonal matters are the only consideration. All other devices must be considered as compromise measures.

Desirable as the correct attack caused by the chiffing of the pipe undoubtedly is in the playing of rapid polyphonic passages, where it contributes no only to the charm of the sound but also to the clarity of the melodic line and to the independence of the voices, it is equally true that this type of attack is not as desirable in the playing of slow legato music. In nearly all other musical instruments the nature of the attack of the tone is variable and controlled by the player. Without this control an expressive device of considerable importance is lost. If for example, the violinist could play only louder or softer, and faster or slower, without any choice as to the vigor and manner of his bowing, or if the flutist could not alter his method of tonguing with the nature of the passage to be played, surely there would be unanimous agreement that these instrumentalists would be under a considerable disadvantage and probably without patronage. It is therefore reasonable to assume that the organist should be deprived of such expressive devices, and that a really musical effect can be produced in the organ when only one type of attach is possible? Is it not true also that the smaller the instrument the more important it is that the organist have such control? This control can be achieved only through the application of mechanical key action which enables the player to effectively influence the attack through the speed with which the valves are opened.

Light and responsive mechanical action is possible and is not to be confused with the clumsy and artistically indifferent mechanism known as the tracker organ. When such mechanisms are applied to modern slide chests which are completely free from the defects mentioned by Mr. Willis in the Autumn Issue, we have a really expressive and ideal organ which we can with some degree of justification and honesty call a musical instrument. Prudence dictates that instruments of this type should not exceed certain limits in size, eight to ten stops per division being maximum. Such organs are possible, and indeed are being built, although the newest and improved models of the Rieger are the only instruments of this kind readily available in the United States today.