100 GIULIANI AND GAGGIO v. ITALY JUDGMENT – SEPARATE OPINIONS 11. We would observe that the authorities had a number of elements available to them in calculating the trajectory of the fatal bullet: the various forensic medical and ballistics reports; the fact that the bullet had fragmented; the fact that an object is shown on film disintegrating in the air shortly before Carlo Giuliani fell to the ground; the theory advanced by the applicants' experts according to which the fragmentation of the bullet could have been caused by factors other than collision with a stone; and the photographs taken shortly before and shortly after the fatal shot and during the autopsy. 12. The photograph taken a few moments before the shot shows the gun positioned at chest height (see also point 6 of Judge Bratza's partly dissenting opinion, annexed to the Chamber judgment), at an angle compatible with the wound sustained by Carlo Giuliani (according to the autopsy report, the bullet entered the body through the left eye socket and exited through the back of the skull, travelling through the body in a downward direction). Accordingly, although it is not impossible, it is unlikely that (a) M.P. raised his gun just as he fired the shot; (b) the bullet ricocheted off a flying object; (c) the angle of collision between the object and the bullet was such as to make the bullet strike the victim very close to where it would have struck him had the gun not changed position. 13. As regards scenario (b) above, it should be noted that the photographs taken just before the fatal shot do not show any stone or other object hovering in the air. This would seem to indicate that in the moments surrounding the firing of the shots the demonstrators were not engaged in intensive throwing of missiles. That suggests that the statistical probability of any of the three scenarios having occurred is low; the likelihood of all three occurring in rapid succession is smaller still. 14. In terms of the Court's case-law, when an applicant adduces prima facie evidence that excessive use was made of lethal force, the onus is on the Government to prove otherwise (see Toğcu v. Turkey, no. 27601/95, § 95, 31 May 2005, and Akkum and Others v. Turkey, no. 21894/93, § 211, ECHR 2005-II). We believe that the same should apply where the Government rely on a statistically unlikely theory in order to counter the applicants' version of events, which is corroborated by visual evidence; it was for the authorities to prove that the very rare events which they alleged actually occurred. However, no such proof was furnished either at domestic level or before the Court. In her decision to discontinue the proceedings the investigating judge herself observed that the ballistics tests had not succeeded in establishing the initial trajectory of the shot. 15. Finally, it seems to us that the Grand Chamber judgment does not place the events giving rise to this tragic case in their proper context. Proceeding as though the case concerned a situation of individual violence, the Grand Chamber concludes that the use of lethal force was necessary in defence of the person concerned under Article 2 § 2 (a) of the Convention