I dissent from the reversal as to appellant McGreen.
The majority reasons (1) that simple assault is a necessarily included offense of robbery; (2) that the allegation that McGreen used a firearm in the commission of the robbery is a part of the robbery charge; and (3) that the “use” allegation is part of the “specific language of the accusatory pleading.” The majority concludes that the charge of robbery with a use allegation encompasses the offense of assault with a deadly weapon, and that the trial court erred in failing to instruct sua sponte on this necessarily included offense. The majority expressly disagrees with People v. Orr (1974) 43 Cal.App.3d 666 [117 Cal.Rptr. 738] and its progeny, which hold that a firearm use allegation under Penal Code section 12022.5 is not to be considered in determining whether the accusation includes a lesser included offense as pleaded. (People v. Benjamin (1975) 52 Cal.App.3d 63, 71-72 [124 Cal.Rptr. 799]; People v. Wilson (1976) 62 Cal.App.3d 370, 374 [132 Cal.Rptr. 813]; People v. Salas (1978) 77 Cal.App.3d 600, 607 [143 Cal.Rptr. 755]; People v. Cole (1979) 94 Cal.App.3d 854, 861-862 [155 Cal.Rptr. 892].)
The majority has articulated a thoughtful and compelling criticism of Orr and its descendants. The Orr court characterizes its holding as a “logical consequence” of the rule in People v. Henry (1970) 14 Cal.App.3d 89 [91 Cal.Rptr. 841]. (Orr, supra, 43 Cal.App.3d at p. 673.) However, as the majority points out, Henry does not provide a solid logical foundation for the holding of the Orr court. A lesser offense is “necessarily included” if it is within the offense specifically and factually charged in the accusatory pleading; the rationale for that rule is that such a pleading thereby puts the defendant on notice that he must be prepared to defend against evidence showing elements of the lesser offense. (People v. Marshall (1957) 48 Cal.2d 394, 405 [309 P.2d 456].) Neither Orr nor cases following it discuss the applicability of that rationale to a pleading charging an offense and a use allegation, and such a pleading unquestionably informs a defendant that he must defend against evidence as to all allegations raised therein.
Nevertheless, I disagree with the majority’s conclusion that it was error for this trial court not to instruct on its own motion on assault with a deadly weapon.
The principles governing the duty to instruct sua sponte in a criminal case have often been repeated. Even in the absence of a request, the trial court must instruct on the general principles of law relevant to the *528issues raised by the evidence. (People v. Sedeno (1974) 10 Cal.3d 703, 715 [112 Cal.Rptr. 1, 518 P.2d 913].) “‘The most rational interpretation of the phrase “general principles of law governing the case” would seem to be as those principles of law commonly or closely and openly connected with the facts of the case before the court.’ [Citations.]” {People v. Flannel (1979) 25 Cal.3d 668, 681 [160 Cal.Rptr. 84, 603 P.2d 1], italics in original.)
The sua sponte rule is designed to “promote the ends of justice by providing some judicial safeguards for defendants from the possible vagaries of ineptness of counsel under the adversary system. Yet the trial court cannot be required to anticipate every possible theory that may fit the facts of the case before it and instruct the jury accordingly.” {Flannel, supra, 25 Cal.3d at p. 683.) Finally, “the duty of the trial court involves percipience—not omniscience.” {People v. Cram (1970) 12 Cal.App.3d 37, 41 [90 CaLRptr. 393].)
In Flannel a defendant argued that the trial court erred in failing to instruct the jury sua sponte that his honest but unreasonable belief that he must defend himself from deadly attack negates malice, so that the offense is reduced from murder to manslaughter. The Supreme Court concluded that although that unreasonable belief doctrine was an extant legal doctrine, it was unique, seldom applicable, and had never been given full substantive discussion by courts. The court held that in cases not yet tried, a sua sponte instruction on that doctrine would be necessary when other sua sponte instruction requirements were met. However, the court concluded that prior to its decision, this inadequately developed doctrine did not express a “general principle of law” presented by the evidence. Therefore, the trial court’s failure to give an instruction on that principle was not error.
A conclusion that the trial court did not err in this case seems equally warranted, if not more so. In Flannel the principle in question was at least an existing legal doctrine, albeit one never given full discussion. Here, however, an unbroken line of authority, including at least one case in which hearing was denied by the Supreme Court (People v. Salas, supra, 77 Cal.App.3d 600), declared that a use allegation was not to be considered in determining necessarily included offenses as pleaded. While one Court of Appeal may decline to follow the prior decisions of another district or division (see McGlothlen v. Department of Motor Vehicles (1977) 71 Cal.App.3d 1005, 1017 [140 Cal.Rptr. 168]), trial courts are bound by decisions of the Courts of Appeal. (Auto Equity *529Sales, Inc. v. Superior Court (1962) 57 Cal.2d 450, 455 [20 Cal.Rptr. 321, 369 P.2d 937].) Furthermore, although the Supreme Court’s denial of hearing is not to be regarded as expressing approval of the propositions of law set forth in a Court of Appeal opinion, “it does not follow that such denial is without significance as to the views of the members of that court.” (McGlothlen, supra, 71 Cal.App.3d at p. 1017, citing DiGenova v. State Board of Education (1962) 57 Cal.2d 167, 178 [18 Cal.Rptr. 369, 367 P.2d 865], and 6 Witkin, Cal. Procedure (2d ed. 1971) Appeal, §§ 669-670, pp. 4581-4584.) Accordingly, until today’s decision, an assault with a deadly weapon instruction because of a use allegation would be improper. As a consequence, even if the majority’s criticism of Orr and the cases which follow it is correct, I cannot agree that the rule which the majority draws from its analysis was a “general principle of law” triggering the trial court’s sua sponte obligation.
Given the majority’s conclusion that the court erred in this case, trial courts will now be faced with conflicting rules as to their obligations under similar circumstances. Not percipience, but prescience will be required, until the matter is definitively resolved by the Supreme Court. The better practice under these circumstances is for the Court of Appeal to follow the established rule, set forth its criticism, and invite the Supreme Court to resolve the conflict. Hopefully, the Supreme Court would, if it followed the majority’s rationale, make its rule prospective to actions not yet tried.
I also disagree with the majority’s conclusion that the prosecutor’s questions and insinuations regarding perjury amounted to prejudicial misconduct.
The testimony of the expert consumes almost 200 pages: he was examined, voir dired, cross-examined, reexamined, and recrossed twice. During this testimony, he (1) admitted that during an earlier trial, he gave his age as 32, corrected himself after realizing he had miscalculated, but then miscalculated again before finally accurately stating that he was 36; (2) explained that confusion as an attempt to clarify an exchange at yet another prior trial, then admitted this explanation was incorrect; (3) had difficulty calculating the number of ounces in a fifth of vodka; (4) admitted he “embellished” his credentials, both at a prior trial and at this trial (by which he meant he presented them in “a pleasing way”); (5) admitted he had testified inconsistently in this case and another as to whether Valium was a tranquilizer per se; (6) misspelled Valium; (7) testified variously that he *530didn’t know how many times he’d talked to appellant’s counsel prior to trial, and that they had talked about 10 times.* Appellant’s own counsel recognized the potentially devastating impact of all of the above on the credibility of this witness, and in his closing argument attempted to minimize these details as “collateral trees.” Despite counsel’s effort to isolate these details as merely collateral, they unquestionably must have impaired the believability and effectiveness of all of this witness’ testimony.
Moreover, certain of his substantive testimony was less than persuasive. For example, after considerable testimony about the significance of various levels of blood alcohol, “zones of unconsciousness,” and the synergistic effect of alcohol and Valium, he was unable to satisfactorily respond to a hypothetical question by the prosecutor as to how a shorter alcohol consumption time would alter the blood alcohol level of an individual of appellant’s size. He then could not, or would not answer the most rudimentary questions: “Q.... Is the difference between 12:00 and 2:30 two and a half hours? That is all you have to answer. Yes or no. A. I don’t know. Q. Can you count it on your fingers? 12:00 to 1:00 is one hour; is that fair to say? A. I don’t have an opinion on that question.” Describing this last exchange in his closing argument, appellant’s own counsel told the jury that there was a simple answer to the hypothetical question, but that instead “I saw a man get all confused, apparently so upset at something that he wasn’t even thinking clearly.”
I would conclude that this witness was incredible, not because of the prosecutor’s objectionable remarks, but because of the expert’s own admittedly inconsistent, inaccurate, and sometimes confusing testimony.
Finally, the Brigham error, standing alone, was not reversible error in this case.
I would affirm appellant McGreen’s conviction.
Petitions for a rehearing were denied July 25, 1980. Scott, J., was of the opinion that the petitions should be granted. The petitions of appellant Race and respondent for a hearing by the Supreme Court were denied August 21, 1980. Clark, J., was of the opinion that the petitions should be granted.
While mathematical miscalculations might seem inconsequential, this witness was presented as an experienced forensic toxicologist and criminalist. He testified that among his special skills were those analytic skills involved in “forensic alcohol,” which he described as the determination of blood alcohol levels and their effect, and courtroom work explaining such data. His arithmetic difficulties loom larger in this context.