I dissent. I cannot conclude that the record contains sufficient substantial evidence to uphold the conviction on count 4—the count involving the Emerson prescription: there simply was no evidence that the prescription was other than genuine. (Cf., People v. Kurland (1973) 33 Cal.App.3d 197, 200-201 [108 Cal.Rptr. 874, 117 Cal.Rptr. 216].)
The majority bases its conclusion that the evidence was sufficient on three bits of circumstantial evidence: (1) the discovery in Jackson’s car of *564“stolen” items apparently belonging to an S. Y. Emerson; (2) the extrajudicial admissions of Hattaway; and (3) the cooperation between Hattaway and Jackson who had just engaged in a transaction involving prescriptions which were proved to have been forged. None of these, considered alone or in the aggregate, constitutes substantial evidence that the Emerson prescription was false.
1. There was no evidence that the Emerson passport, credit cards and checks were stolen. Indeed, the prosecutor conceded that fact when the receiving stolen property charge based on those documents was dismissed.
2. While Hattaway’s extrajudicial admissions implicated him in some sort of nefarious dealings with Jackson concerning the drugs that had been ordered by way of the Emerson prescription, absent bald speculation that fact in no way proves that the Emerson prescription was false.
3. The prosecution presented a handwriting expert who testified that the two prescriptions which had been personally passed by Jackson were forgeries. No comparable evidence regarding the Emerson prescription was presented. While the interaction between Hattaway and Jackson may have raised a strong suspicion that the third prescription was also forged, such suspicion cannot substitute for solid proof of guilt. I would reverse count 4.
I also have my doubts on the Aranda-Bruton point. The majority holds that there was no Aranda-Bruton error because there was substantial evidence linking Jackson with the crime other than Hattaway’s extrajudicial admissions. In support of this proposition, the majority cites People v. Epps (1973) 34 Cal.App.3d 146 [109 Cal.Rptr. 733] and People v. Romo (1975) 47 Cal.App.3d 976 [121 Cal.Rptr. 684]. From my reading of Epps and Romo—two doubtful cases—the majority overstates their effect. In Epps, the extrajudicial statements showed only that the codefendants were acquainted with each other. In this they were cumulative. Further they did not directly implicate the codefendant who had not made the admissions in the crime. In Romo, the principle adverted to is merely a dictum.
The situation here is vastly different from that in Epps. Hattaway’s admissions clearly and directly implicated Jackson in the charged crimes. Because the court denied Jackson’s alternative motion to sever or strike *565the parts of the statements which implicated him, and because Hattaway did not testify, there was Aranda-Bruton error. The other evidence of guilt may have rendered the error harmless—the present posture of the case makes it unnecessary for me to decide that question—but it did not make the error nonexistent.