This will construction ease turns on whether extrinsic evidence is admissible to show the testator’s intent. At issue is a bequest “to my nieces and nephews.” If the challenged evidence is admissible it conclusively shows the testator intended to include nieces and nephews by marriage— *480that is, those who were the children of brothers and sisters of the testator’s deceased wife. We think a latent ambiguity exists and that the evidence is admissible to resolve it. Accordingly, we reverse a trial court determination to the contrary.
The testator had one blood niece and one blood nephew. His deceased wife had nineteen blood nieces and nephews. According to the challenged evidence the testator told his lawyer he wanted his estate to go to his twenty-one nieces and nephews, two on his side of the family and nineteen to whom he was related through his deceased wife. The testator at first sought to specify the twenty-one beneficiaries by name. He was, however, unable to supply all the names and addresses. The residual clause, when drawn, merely specified that the property pass “to my nieces and nephews, who survive me, share and share alike.”
The challenged extrinsic evidence went further. It showed that, although there was certainly no estrangement, the testator was only casually associated with the niece and nephew with whom he was related by blood. The testator had not seen either of them for at least ten years prior to his death. He did not have the address of the blood nephew and did not even know whether the blood niece was alive.
On the other hand the testator had a close and affectionate association with various of the nieces and nephews, with whom he was related through his wife. These relationships continued after the death of the testator’s wife, some five years prior to his own.
The blood nephew and niece brought this declaratory judgment action, asking that the bequest be construed to exclude the nineteen nieces and nephews whose relationship with the testator was not by blood.
The trial court found the bequest to be unambiguous and held that, according to its common legal definition, the term “nieces and nephews” referred only to those related by blood. Hence none of the extrinsic evidence was admitted to show the testator’s intent to the contrary. As a result the trial court held that the blood niece and blood nephew were the sole beneficiaries under the will. This appeal followed.
I.
A declaratory judgment to construe or interpret a decedent’s will is tried in equity and our review is de novo.
Russell v. Johnston, 327 N.W.2d 226, 228 (Iowa 1982). In a de novo review, of course, we make findings of fact anew. In so doing, we give weight to the fact findings of the trial court but we are not bound by them. Iowa R.App.P. 14(f)(7).
There is no quarrel about the fundamental precepts which apply in will interpretation cases:
It is well settled law (1) the testator’s intent is the polestar and must prevail;
(2) his intent must be gathered from a consideration of (a) all the language contained in the four corners of his will, (b) his scheme of distribution, (c) the circumstances surrounding him at the time he made his will and (d) the existing facts; and (3) technical rules or canons of construction should be resorted to only if the language of the will is ambiguous or conflicting or the testator’s intent is for any reason uncertain.
In re Estate of Larson, 256 Iowa 1392, 1395, 131 N.W.2d 503, 504-05 (1964).
Neither is there any challenge to the trial court’s determination that there is no patent ambiguity in the bequest and that the term “nieces and nephews” commonly refers only to those related by blood. See Shalkhauser v. Beach, 14 Ohio Misc. 1, 233 N.E.2d 527, 530 (1968); see also 80 Am.Jur.2d Wills § 1198 (1975); 95 C.J.S. Wills § 667 (1957).
II. What is disputed is the trial court’s failure to find a latent ambiguity. The appellants contend a latent ambiguity surfaced when the language of the bequest did not square with the actual number of blood nieces and nephews. That is, the bequest, as to both genders, was plural whereas in point of fact there was but one blood niece and but one blood nephew.
*481The trial court dismissed this anomaly on the basis of In re Weber’s Estate, 155 Pa.Super. 403, 38 A.2d 362, 363 (1944). It viewed the bequest as merely one to a class and saw no ambiguity in the fact that only two members of the class were found to exist.
We recently considered a claim of latent ambiguity in a will in Estate of Kiel, 357 N.W.2d 628 (Iowa 1984). In Kiel, we again pointed out the necessity for a patent or latent ambiguity before extrinsic evidence can be admitted to show a testator’s intent. 357 N.W.2d at 630.
The nineteen appellants insist, and we agree, that the wording of the bequest simply did not apply to a singular niece and a singular nephew. The term is to be read in context. The second paragraph of the will provided:
Subject to the provisions of clause FIRST, I hereby give, devise and bequeath all of my estate and all the property of which I may die seised and possessed, and to which I may be entitled at the time of my death, and which I may acquire after my death, of whatsoever kind and nature, and wheresoever situated, whether real, personal, or mixed, to my nieces and nephews, who survive me, in equal shares. If one or more of my nieces and nephews should predecease me, then I devise that deceased niece or nephew’s share to my surviving nieces and nephews, in equal shares.
The nineteen appellants argue that it makes no sense to describe plural nieces and plural nephews if only one of each existed. They think the inappropriateness of the plural terms in describing singular persons becomes more pronounced when the last sentence is considered: “If one or more of my nieces and nephews should predecease me, .... ” For many years the testator had at most but one of each who was related to him by blood.
The blood niece and blood nephew counter by pointing out that the language described nothing at variance with a routine class gift. But we think the suit cannot be resolved by merely defining the bequest as a class ■ gift. Under either interpretation proposed here a class gift is created. The dispute is over how large the class is. In resolving that dispute we agree with the nineteen appellants when they insist that the words used make no sense if they are applied to describe a class of one niece and one nephew.
We do not suggest that, in a class gift to nieces and nephews, the class could never be reduced to a singular niece or nephew. Neither do we hold that a bequest of the kind involved here could not be shown to apply to a single niece and nephew. We merely hold that, on the facts here, a latent ambiguity appeared when it developed that the testator had but one blood niece and one blood nephew.
Our cases on testamentary construction have tended more and more to seek out the testator’s intent. To be sure, we are bound by the unambiguous language of a will. In re Estate of Kiel, 357 N.W.2d at 630. On the other hand, we have a growing disinclination to allow a testator’s intent to be frustrated by a technical but obvious mistake in draftsmanship. Compare Estate of Lepley, 235 Iowa 664, 670-71, 17 N.W.2d 526, 529 (1945) with Elliott v. Hiddleson, 303 N.W.2d 140, 141-42 (Iowa 1981). We hold that the extrinsic evidence should have been admitted to resolve the latent ambiguity we have described.
On our de novo review of all the evidence we find it was the testator’s intent to bequeath his property to his surviving nieces and nephews and to his deceased wife’s surviving nieces and nephews, share and share alike.
The judgment of the trial court is reversed and the case is remanded for entry of a judgment in conformity with this opinion.
REVERSED AND REMANDED.