Snow, J.

The principal question involved is the ascertainment of the quantity of water to the use of which the defendants are entitled under the terms of the Barker grant. The defendants claim that the measure of their rights under the deed is the quantity of water which the fore-bay and open flume would carry. They concede that the quantity of water granted is specified in the deed, namely, the amount necessary "to run and carry three run of stones, ■a corn cracker and smut mill,” but say that by necessary implication it was the quantity of water so required when applied to the old wooden wheel then in the mill or to wheels of that type; that the restriction in the width of the flume was also intended as a limitation of the quantity of water granted; that the burden of proving the quantity of water necessary to run the specified machinery applied to the old wheel, or to wheels of that type, was upon the plaintiff; that, the plaintiff having failed in its burden of proof, and the court having found that without the installation of more modern wheels the grantees could not avail themselves of sufficient water to run the specified machinery, therefore the defendants are entitled to a decree based upon the capacity of the flume; that the language of the deed as thus construed is plain and unambiguous and that, therefore, it was error for the court to receive evidence of the practical construction of the deed by the parties.

In this argument the defendants lose sight of the fact that the finding of the court, upon which they rely as an essential premise, is based upon evidence dehors the deed. If extraneous evidence is ad-, missible to aid in the construction of the deed no reason is perceived for limiting such evidence to that relating to the incapacity of the old wheel, to the exclusion of all other competent extrinsic evidence bearing upon the intention of the parties.

However, the defendants’ contention that the provision "the width of said flume and rack is not to be in any manner increased” was intended as a measure of the quantity of water, is clearly untenable. The quantity is expressly limited to the requirements of "three run of stones, corn cracker and smut mill.” The restriction *145 upon the width of the flume appears in the clause providing for repairs. Its presence there is abundantly accounted for by the evidence that a widening of the flume would endanger the safety of the dam. That it was not intended as a measure of the water is further evident from the fact that no limitation is placed upon either the depth of the flume or the velocity with which the water was to be drawn, each of which elements is quite as essential to the determination of the water discharge as is the width of the flume. This conclusion is confirmed by the further finding that it was not then customary to build open top flumes carrying only enough water to supply the wheels, as under such conditions the water would get “stirred up” and a quiet state of the water in such a flume constitutes an important element in the proper utilization and conservation of the stream.

It is clear that the parties adopted as the measure of the rights granted the quantity of water necessary “to run and carry three run of stones, corn cracker and smut mill.” The wheel is not mentioned in the deed. Its capacity to discharge water was not made the measure of the rights granted either expressly or by necessary implication (Hut chins v. Berry, 75 N. H. 416, 417). It is important, therefore, only as it was a part of the existing appliances in the mill at the date of the deed. Horne v. Hutchins, 71 N. H. 128, 136; Hutchins v. Berry, 73 N. H. 603, 604. If there were evidence of its discharge, and if the power which it generated could be correlated with the power required to “run and carry” the specified machinery, it would be an important piece of evidence. But the only known fact bearing upon the relation of the wheel to such machinery is its incapacity to run and carry it. True, it had been the motive power in the old mill. But whether it could run the two runs of stones and their accessories at one and the same time, or whether it could run only one run of stones at a time, is not found. This situation tends to confirm the court’s finding that the parties to the Barker deed deemed the capacity of the wheel immaterial to the determination of their rights. This finding is a logical inference from the facts that neither of them made any effort to ascertain its capacity or to preserve any facts or data in respect thereto, coupled with the early installation and long use without objection of a different type of wheel. But of whatever value, if any, the wheel might have been in the settlement of the problem presented, its present uselessness as an aid in ascertaining the intention of the parties to the Barker deed is established by the finding that it is now utterly impossible to *146 ascertain how much water was required to run it. But the loss of this evidence, even though it were important, does not render unsolvable the problem of determining the rights of the parties, nor entitle the defendants to the right to use all the water their fore-bay and flume could carry irrespective of the velocity with which the water might be drawn. Other competent evidence, more or less weighty, remains from which the intention of the parties may be ascertained. Smith v. Furbish, 68 N. H. 123; Weed v. Woods, 71 N. H. 581, 583.

Where the language of a deed is susceptible of more than one meaning, evidence of the practical interpretation of the deed by the acts of the parties is admissible. Bell v. Woodward, 46 N. H. 315, 332; Winnipiseogee &c. Company v. Perley, 46 N. H. 83, 103, 108. When the parties to a deed have so acted in relation to any of its provisions as to show their understanding of them, and this course of action has continued for a long time, such conduct is entitled to great weight in determining the question of intent, and will be adopted by the court unless the language of the deed is clearly incapable of such construction. Morrill v. Weeks, 70 N. H. 178, 180; Day v. Towns, 76 N. H. 200, 201, 202, and cases cited. This principle of construction has been held to be particularly applicable in the interpretation of conveyances of water rights. Fuller v. Daniels, 63 N. H. 395, 397; Fowler v. Kent, 71 N. H. 388, 393, 394; Horne v. Hutchins, 71 N. H. 117, 118, 127.

This is not a case of the admission of extrinsic evidence to control a written contract unambiguous in its terms (Cummings v. Blanchard, 67 N. H. 268, 272). It is manifest that the express grant of sufficient water to run and carry three runs of stones, corn cracker and smut mill, under the head and with the appliances in the Lyman Mill in 1875, or with appliances of similar type, is not adequate data from which the rights of the grantee can now be computed and expressed in terms recognized in modern hydraulics, namely, the number of cubic feet of water per second applied under a given head. In this situation evidence of what the language of the grant meant to the parties as shown by their conduct during the years immediately following the grant is clearly competent. Fuller v. Daniels, supra. The court has found that the parties to the deed at the time of the grant contemplated the removal of the old wheel and the installation of more modern wheels in its place; and that they and their successors in title, by their conduct for the thirteen years following the grant, viz., from 1875 to 1888, accepted the discharge of the Buzzell wheels, *147 so installed, as the test by which the rights of the grantees under the deed were to be measured. These findings warrant the adoption by the court of this test as a basis for its decree.

But the defendants except to the finding that both grantor and grantee contemplated the removal of the old wheel and the installation of more modern wheels because they say (1) that there is no such evidence in the case; (2) that such evidence would not have been admissible; (3) that the evidence is to the contrary. The contrary evidence relied upon is the testimony of Willey, the only living party to the deed, to the effect that before they (Hadlock and Willey) bought the Buzzell wheels they had obtained a water wheel from Barre, Vermont, because they had decided to put in another run of stones; that before this wheel was put in operation Mr. Buzzell came along and induced them to buy of him the four wheels which were installed; and that the only knowledge he (Willey) then had of the Buzzell wheels was the fact that there was one in the sawmill on the other side of the river. The importance of Mr. Willey’s testimony is much impaired by his want of memory and by the fact that he was not the practical miller of the firm. It is not perceived, however, in what respect his testimony is even contradictory of the court’s finding. It is to be noted that the court has not found that the parties contemplated installing Buzzell wheels. Furthermore, if the evidence could be construed as contradictory it would not avail the defendants. The contention that the evidence conclusively disproved that the parties contemplated the removal of the old wheel and the installation of more modern wheels cannot be seriously considered.

The defendants’ double difficulty as respects the presence and admissibility of this evidence arises (1) from their misconstruction of the word “contemplated” as used in the finding and (2) the use to which the finding is put. To contemplate is to have under consideration as possible or probable. The word as used by the court is not a finding of the intention of the parties in contradiction of the express terms of the deed (Electric Light Co. v. Jones, 75 N. H. 172, 174) but a finding, from the condition of the premises and the subsequent conduct of the parties, of their probable purpose in doing what they did. Nothing is said in the deed as to the manner in which the power to run the specified machinery was to be generated. From the fact of their purchase of a valuable water right it is a reasonable inference that they expected to make some use of it. The fact that the old wheel was inadequate to propel the designated *148 machinery and that wheels of that type were no longer being installed, leads to the conclusion that they had in mind the making of some change in the motive power. The fact that the old wheel was soon removed and more modern wheels installed by the grantees, with the knowledge of the grantors, is some evidence that at the date of the deed the parties were considering the substitution of new and more modern wheels for the antiquated wheel. The fact that both parties accepted and for many years acquiesced in the resulting draft of water from the common source tends to show that what was done was pursuant to the earlier contemplation of the parties. Oakland Woolen Company v. Company, 101 Maine 198. See Blake v. Madigan, 65 Maine, 522, 528. The finding was based upon competent evidence and was itself competent proof in aid of the construction of the language of the deed. This was the only purpose for which it was used.

In applying the test adopted the court found that the Buzzell wheels, as installed, required in the aggregate 105.1 cubic feet of water per second to produce their maximum power. This finding was a mathematical deduction depending upon a subsidiary finding that when installed these wheels carried a head of 6.1 feet above the horizontal center of their gate openings. The defendants except to the latter finding on the ground that there, was no evidence to sustain it. It is conceded that the difference in the elevation of the dam in 1875 and that of the sill supporting the floor on which the Burn-ham wheels were set in 1888, according to a plan made in June, 1921, is 10.2 feet. The finding of 6.1 feet effective head as applied to the Buzzell wheels was obtained by first deducting 1.5 feet, measuring the supposed difference in the elevation of the Buzzell above that of the Burnham wheels. As the Buzzell wheels received the water through an aperture 30 inches high, the bottom of which entered the wheel casing at the level of the top of the wheel, a further deduction of 29 inches was allowed, of which 14 inches represented the height of the wheel and 15 inches the further height of the center of the aperture. The 29 inches was called 2.4 feet. Deducting the sum of 1.5 feet and 2.4 feet from 10.2 feet leaves a difference of 6.3 feet net head. The expert hydraulic engineer who made the computation made the net head 6.1 feet. It is not apparent whether this difference represented error or whether the witness made an allowance for the supposed thickness of the plank upon which the evidence shows the Burnham wheel set above its sills. As the difference may have been accounted for upon the evidence no question *149 of law is raised. The alleged error in the computation upon which the defendants principally rely, however, is in the assumption of the expert witness that the elevation of the Buzzell wheels was 1.5 feet above that of the Burnham wheels. Of this they say there was no evidence. McFarland, who operated the Buzzell wheels from 1880 to 1888 and who in the latter year, as owner, built the Granite State Mill and installed the Burnham wheels, testified that the Buzzell wheels “did not set down to the water.” In answer to an interrogatory as to “his best recollection how high up” he testified, “I don’t know, somewhere from one to two feet. I should think that those wheels set up from the bench perhaps not more than one foot. I could not tell. I know those Buzzell wheels did n’t set down level with the water and when I put in the new [Burnham] wheels I calculated to set them level with the ordinary water.” Counsel for the plaintiff interpreted this answer as meaning that the elevation of the Buzzell wheels was from one to two feet above that of the Burnham wheels, and so assumed in his questions to the expert calling for his computation of the effective head applicable to the former. The defendants say that the correct interpretation of this testimony was that the Buzzell wheels were set one to two feet above the tail water, and, as the Burnham wheels on the plaintiff’s evidence set 1.55 feet above tail water, that the effective head applied to the Buzzell wheels should, therefore, have been found 7.84 feet instead of 6.1 feet. The language of the witness would bear either construction. As the interpretation of the testimony of a witness in the absence of a plain mistake is for the trial court it must be assumed that the court adopted that interpretation which is consistent with its finding. The defendants’ exception therefore raises no question of law.

The defendants except to the decree so far as it limits their draft of water to “the ten working hours of each day of the year, viz. from seven to twelve in the forenoon and from one to six in the afternoon”; also to the refusal of the court to rule that they are entitled to use the water for the full period of twenty-four hours each day.

Where from the language of a grant of water power the parties have left it doubtful whether their intention is to limit the purpose for which the water may be used or only to prescribe the quantity which may be drawn the latter construction will be favored. Johnson v. Rand, 6 N. H. 22, 23; Dewey v. Williams, 40 N. H. 222, 228; Fowler v. Kent, 71 N. H. 388, 394; Hutchins v. Berry, 74 N. H. 225, *150 229, and cases cited. Under these authorities the grant of the “grist mill . . . together with . . . the right to use sufficient water to run and carry three run of stones, corn cracker and smut mill,” in the absence of controlling evidence to the contrary, must be construed as limiting the volume of water that may be drawn and not the purpose or manner of its use. We find in the reported facts no sufficient evidence to overcome the natural inference in this regard. The quantity necessary to “run and carry” the specified machinery having been found upon competent evidence to be 105.1 cubic feet per second, the deed is to be read as though these figures had been written into the grant.

The grqnt is of the right to take this prescribed volume of water whenever the grantee wishes to do so unless there are surrounding circumstances which disclose a different intention. Carleton Mills Company v. Silver, 82 Maine, 215, 221; Oakland Woolen Company v. Union Gas & Electric Company, 101 Maine 198; Wilton Woolen Company v. G. H. Bass & Company, 112 Maine, 483; 67 L. R. A. 397; 27 R. C. L. 146. The deed is silent as to the'hours during which the specified quantity can be drawn unless the mere fact that the parties made the gristmill use a measure of the right granted can be treated as limiting the time to the customary hours of carrying on that business. Except as it may be a matter of common knowledge, the sole evidence of the customary hours of running gristmills, as well as of the customary hours of operating this mill prior to the grant, consists of the testimony of witnesses as to the practice at the mill following the date of the grant. From this it appears that there was no fixed rule limiting the hours of running, but that, on the other hand, the time of operation varied with the seasons and was controlled solely by the exigencies of the business. Business was good in the fall, winter and spring and slack in summer. While it could be found that the average daily use for the year was about ten hours it appears that on some days it was more and on other days less, and that occasionally the mill was run evenings and nights. This evidence tends to disprove rather than to prove the existence of a custom to operate the mill according to a definite schedule. Apparently the grantees and their successors in title have used the water at will, as they have had occasion. Treated as evidence of the manner or custom of operating this mill or mills of its character prior to the grant, it affords an insufficient basis for a finding of an intention to limit the hours of running in accordance therewith. Nor is it important, treated as evidence of practical construction by *151 the parties. The fact that the water was not used for twenty-four hours daily is of no significance, for up to the present time there has been no occasion for such use. It might have been otherwise if there had been some customary forbearance, some shutting down of the mill when there was need for operation.

Nor is there anything in the situation of the property to warrant the inference that the parties had in mind any limitation in the hours of operation. Had the grant been of a right to draw from a reservoir upon which the mills were dependent so that duration of time was an important factor in the effect of the use upon the remaining interest of the grantor, there would be some reason for concluding that the parties understood that the use was limited to ordinary working hours. But when, as here, there is substantially no reservoir and the grant is practically of a portion of the run of the stream, there is nothing in the situation of the property upon which to base an inference that the parties understood the grant to be limited as to time of use.

Furthermore the parties must have intended a reasonable use of the right granted. It was conceded by the plaintiff in argument that defendants’ use of the water for twenty-four hours a day would not be a material detriment to the plaintiff. It is reasonable to suppose that the same situation existed at the time of the grant. Evidence that a certain use is beneficial to the grantees and not detrimental to the grantor tends to prove that a grant of such a use was intended. Such a use is therefore reasonable under the facts of the particular case; and being a reasonable use the grantee is entitled to exercise it. Gardner v. Webster, 64 N. H. 520, 521, and cases cited; Boston &c. Railroad v. Railroad, 65 N. H. 393, 465; Wilcox v. Kendall, 63 N. H. 609; Jewell v. Clement, 69 N. H. 133, 134; Gill v. Ferrin, 71 N. H. 421, 423; Kendall v. Green, 67 N. H. 557, 563.

The defendants’ exceptions to the decree, and to the court’s rulings, so far as they limit the hours of draft of the specified quantity of water, are sustained.

When the dam was repaired and raised two feet in 1884, McFarland, the owner and occupant of the gristmill, contributed one-eighth of the expense thereof. The defendants except to the provision in the decree that “the defendants have acquired no rights to additional water by the raising of the dam.” It appears to be the defendants’ position that McFarland’s contribution ipso facto invested him with the title to a quantity of water, which they term “one eighth of the extra water” in addition to that covered by the grant. *152 TMs claim is supported neither in law nor in fact. It is clear that the raising of the dam did not increase the quantity of water flowing in the stream. Its sole function must have been to make the water of the stream more available to the mills on either side of the stream by increasing, the head at each mill and by providing additional storage in times of low water. As respects the head it was mutually beneficial since each party could develop more power with a given quantity of water. If the resulting conservation of water incidentally enured to the greater advantage of the sawmill owner’s inferior privilege it did not, in the absence of an agreement, effect a transfer to McFarland of a portion of the sawmill owner’s title. The clause of the decree containing the particular language excepted to appears to us to state correctly the limit of the defendants’ rights, viz.: “The defendants have acquired by prescription as an easement appurtenant to the mill property, a right to the additional head and reservoir which was created by raising the dam two feet. Aside from the additional head and the enlargement of the reservoir incidental to the increase in the height of the dam the defendants have acquired no right to additional water by the raising of the dam.” As there is no evidence that the gristmill used any more water after than before the dam was raised the decree gives the defendants the full measure of their prescriptive rights, which is their only source of enlarged title.

Exception is taken to the finding of the court that “the trash rack at the head of the open-top flume was made of stock that was an inch thick, and having one-inch opening between each piece” and to the decree that the defendants’ flume must be “equipped with a suitable rack.” Except as to the position of the rack in the flume the finding is abundantly supported by the evidence. The position of the rack would not interfere with the draft of 105.1 cubic feet per second nor diminish the head, and is therefore wholly immaterial to any issue .here. As the Barker deed expressly provides that the “grantees are to keep and maintain the flume and rack of said Grist Mill in good repair” the defendants have here no valid exception to the decree.

There is no merit in the defendants’ exception to the denial of their motion for a decree that they have acquired title by prescription to the use of all the water that the fore-bay, or that the round penstock subsequently installed, would carry. It is sufficient to say that the evidence does not conclusively establish all the requisites to such a title.

Nor is there any merit to the claim that it conclusively appears, *153 contrary to the finding of the court, that the plaintiff has waived its remedy or lost its rights by laches.

Exceptions which have been rendered immaterial by the conclusions here reached, and exceptions which have not been presented in argument, have not been considered. In view of the conclusions reached it is understood that the plaintiff’s exceptions are waived, and they have therefore not been considered.

Decree modified.

All concurred.