This testator made his will in 1878 and the trust which he created *35was to endure as a whole until the death of the survivor of his nine children, and it was inevitable that births and deaths would make many changes in the family. He died in 1885.
Notwithstanding the long term of the trust testator designated his beneficiaries as “children” and the “children” of “children”, which word in its strict sense excludes more remote issue. The auditing judge construed “children” in the sense of issue and admitted great-grandchildren to share in the distribution of principal. Exceptions have been filed by the grandchildren.
The will provides as follows:
“Upon the death or marriage of my wife the whole income of my Estate shall be equally divided among my children during their respective lives, but should any of my children die leaving a child or children then such child or children shall take such portion of the income of my Estate as the parent of such child or children would have taken if living, but should any of my children die leaving no children then the portion of the income of my Estate given to such child shall be equally divided among the survivors of my children and the child or children of those who are dead, the child or children of a deceased child or children to take by right of representation as above provided. . . .
“After the death of all my children, then my Estate shall be equally divided among their living children— per stirpes — that is they shall take by representation of their parents respectively, and not per capita.”
There were two adjudications during the continuance of the trust in favor of the right of great-grandchildren to share in the income. These rulings were based upon the general intent of testator to send the income down in the line of the blood. No exceptions were taken to these adjudications, but as they related only to income during the continuance of the trust, they are not binding in the final distribution of principal.
*36The history of the case shows that testator was survived by all of his nine children. There were also nine grandchildren living at the date of the will, twelve at the date of his death, and six others were born thereafter.
Now at the termination of the trust, nine grandchildren survive who are the children of three deceased children.
Two children of testator left no children but grandchildren, the great-grandchildren of the testator. These were awarded parents’ shares by the auditing judge.
Four children of testator left no issue.
Exceptants would reduce the stirps from five to three, and in two of the three would exclude great-grandchildren from sharing with their uncles and aunts. If the exceptions are sustained, great-grandchildren whose total interest under the adjudication is 52 percent will be disinherited.
In trusts of long duration the disinheritance of lineal descendants may sometimes be avoided by a construction that remainders are vested in deceased parents: Ross v. Drake, 37 Pa. 373, 377; Nass’ Est., 320 Pa. 380. This remainder, however, is contingent. It is given to “living” beneficiaries and the estates of the dead are eliminated. The great-grandchildren are disinherited unless an extended meaning can be given to the word “children”.
The general rule long settled in Pennsylvania is stated in Hallowell v. Phipps, 2 Wharton 376, 379, as follows:
“Under a bequest to children, grandchildren and other remote issue are excluded, unless it be the apparent intention of the testator, disclosed by his will, to provide for the children of a deceased child. But such construction can only arise, from a clear intention or necessary implication; as where there are not other children than grandchildren, or when the term ‘children’, is further explained by a limitation over, in *37default of issue. The word ‘children’ does not ordinarily, and properly speaking, comprehend grandchildren, or issue generally. Their being included in that term, is only permitted in two cases, viz., from necessity, which occurs when the will would remain inoperative, unless the sense of the word, ‘children’ were extended beyond its natural import; and where the testator has clearly shown by other words that he did not intend to use the term ‘children’, in the proper actual meaning, but in a more extensive sense.”
Other cases frequently cited in support of the rule are Hunt’s Estate, 133 Pa. 260; Page’s Estate, 227 Pa. 288; Williamson’s Estate, 82 Pa. Superior Ct. 444; Worstall’s Estate, 125 Pa. Superior Ct. 133.
The well-recognized exceptions, most frequently applied, are where there is a remainder to “children” followed by a gift over in default of “issue” (Albertson’s Estate, 329 Pa. 372), and where the testator makes use of the words “heirs”, “issue” and “children” indiscriminately: Ball v. Weightman et al., 273 Pa. 120; Joyce’s Estate, 273 Pa. 404; Ziegler Estate, 356 Pa. 93. These are regarded as clear indications that testator did not intend to use the word “children” in its actual meaning, but in the more extensive sense of “issue”.
In Campbell’s Estate, 202 Pa. 459, the remainder was to “children” of testator’s daughters, with a gift over should a daughter die without “children”. A daughter died without children but leaving grandchildren. The gift over was construed as an indication of the intent of testator to send shares down in the line of blood per stirpes, and thus include the grandchildren. Judge Penrose in Page’s Estate, 227 Pa. 288, cited supra, distinguished and explained Campbell’s Estate by pointing out that the primary remainder was a vested estate which required the gift over to be construed strictly to prevent a forfeiture.
Campbell’s Estate in its gift over in default of “children” should be compared with Albertson’s Estate, *38supra, and other similar cases where the gift over was in default of “issue”. Campbell’s Estate broadens the effect of a gift over and makes it operate to include grandchildren whether the reference is to “children” or “issue”.
A. L. I. Restatement of the Law of Property, §285, approves the conclusion in Campbell’s Estate in comment (f), in the following words: “The combination of the gift to the children of B and of the gift over upon the failure of the children of B is a factor which tends to establish that the conveyor has used ‘children’ to denote ‘descendants’ since a conveyor is seldom likely to intend the gift over to be effective if the designated parent dies survived by descendants but not survived by children.”
In the instant case there is no gift over in default of “children” insofar as concerns the principal of the trust, but there was such a gift over of income during the pendency of the trust, and the previous adjudications of this court, which awarded the income to grandchildren, were based in large part upon that fact.
The main reliance of the present adjudication is upon Disston Estate, 349 Pa. 129, where the phrase “per stirpes and not per capita”, following a designation of “grandchildren”, conveyed the sense that “grandchildren” was intended to mean “issue”.
Disston Estate, in the emphasis which it gives to the phrase “per stirpes”, is a further abatement and relaxation of the rule excluding remote issue, and, as stated by the auditing judge, is not free from doubt, because the court en banc sustained exceptions over the dissent of the then auditing judge, and the Supreme Court was evenly divided in its affirmance.
The phrase in the Disston will was: “I direct that my estate shall be divided among my grandchildren then living per stirpes and not per capita.”
*39In the instant will it is: “My estate shall be equally divided among their (testator’s children’s) living children — per stirpes — that is they shall take by representation of their parents respectively, and not per capita.”
Prior to Disston Estate, the expression “per stirpes” was not sufficient to extend the meaning of the words “children” and “grandchildren”. It was regarded as fixing the quantum each beneficiary should take, and not as indicating a substitutionary gift to more remote descendants.
In Long’s Estate, 228 Pa. 594, income was given to testator’s children and a grandchild, with a gift over should any of “my children” die without lineal descendants, to other legatees “per stirpes and not per capita”. Upon the death of the grandchild, his share was held vested, and not divested by the gift over of the shares of “my children”. The subsequent direction that the alternative legatees should take “per stirpes” was not of sufficient importance to overcome the plain words of the will.
In Hogg’s Estate, 329 Pa. 163, there was a gift at the termination of a trust to grandchildren “per stirpes”. Certain children were born to testator after the date of his will, and they received intestate shares in partial revocation of the will. The contention was that their children were excluded from the class of grandchildren because “per stirpes” meant a taking by right of representation of a parent, and the parents having already received an intestate share, their children could take no more. This contention was rejected and all grandchildren included, because “per stirpes” was not used in the sense of taking by representation but to indicate a taking by families.
In Burk’s Estate, 21 Dist. R. 357, the gift at the termination of a trust was to “the children of my said sisters per stirpes and not per capita”. It was held that the word “children” did not include grandchildren, *40Judge Gest stating that the words “per stirpes” did not affect the question, “as they simply indicate the proportionate share of the whole estate which is given to the children of each sister as a class”.
In Hummel’s Estate, 49 D. & C. 1, in default of issue of testator, the estate was given to the “children” of testator’s brothers “per stirpes”. Again it was held that the expression “per stirpes” was not sufficient to extend the meaning of the word to include grandchildren. Van Dusen, P. J., dissented, on the ground that the phrase “per stirpes” has a dual significance, first a taking by families in the first generation, and second, substituted gifts to other generations.
In our opinion, the time has come, as in Disston Estate, for a more liberal interpretation of gifts to lineal descendants of testator at the termination of a family trust. We would not disturb the primary meaning of the words “children” and “grandchildren”, but would give effect to any reasonable indication that the words were used in an extended sense. Lineal descendants are the favorites of the law, whatever the rule may be as to remote collaterals.
When testator speaks of his own “children” both layman and lawyer know that he is not thinking of grandchildren. He has intimate knowledge of his own children, and there is a natural distinction between them and grandchildren and more remote lineals.
On the other hand, where there is a gift to grandchildren at a period long after testator’s death, and they are no better known to him than great-grandchildren, there is little reason for discrimination, and a concentration upon part of the family of all the family wealth.
As has been stated, a gift over in default of “children” sufficiently indicates an intention to send the property “down in the line of blood”: Campbell’s Estate, 202 Pa. 459, supra; A. L. I. Restatement of the Law of Property, §285, comment (/).
*41In our opinion, the phrase “per stirpes” shows a similar intent.
“Per stirpes” has a two-fold significance. It may be used to fix the shares of the parties (Hogg’s Estate, 329 Pa. 163), or it may mean a taking by representation: Shoch’s Estate, 271 Pa. 165.
In the instant case, it indicates substituted gifts upon a representative basis, and gives the word “children” the sense of “issue”: Disston Estate, 349 Pa. 129.
This is the rule of A. L. I. Restatement of the Law of Property, §285(2) (d), where it is said in comment (k):
“When the language is unmistakably explicit in directing a distribution in which the offspring of a deceased person represents such deceased person, no one can successfully urge the exclusion of grandchildren or more remote descendants of the designated parent. Such explicitness usually takes the form of either the addition of the phrase ‘per stirpes’ to a limitation in favor of the ‘children of B’, or the inclusion of detailed substitutional dispositions, in terms made applicable to the share of each child (see §302). . . .”
This testator further directs that the children of children “shall take by representation of their parents respectively”, which emphasizes that he had stirpital distribution and the principle of representation in mind.
“Where ambiguous or contradictory expressions appear in a will, the law adheres as closely as possible to the general rules of inheritance”: Windt’s Estate, 110 Pa. Superior Ct. 124.
“The statute of distributions governs in all cases where there is no will; and where there is one, and the testator’s intention is in doubt, the statute is a safe guide”: Ashburner’s Estate, 159 Pa. 545; Sipe’s Estate, 30 Pa. Superior Ct. 145.
*42The disinhersion of lineal descendants in 52 percent of the estate is not consonant with testator’s general plan for a family or stirpital distribution.
The exceptions are dismissed and the adjudication is confirmed absolutely.