Dr. Paul Leeland Kirk, professor of Criminalistics and forensic expert for the Sheppard defense team, stated in his affidavit that he examined the crime scene, various pieces of evidence held by the prosecutor and police, as well as blood samples collected and mailed to him. His investigation began on January 22, 1955 and continued for several months. Dr. Kirk used blood spatter analysis to conclude, among other things, that:
- The murderer was left-handed (Dr. Sam Sheppard was right-handed)
- Injuries to the victim's teeth indicate that the victim had bitten the attacker's hand
- A large bloodstain on the closet door was likely from the attacker's bloody hand
- Testing of the blood stain showed it did not match that of Sam or Marilyn Sheppard, so the attacker must have been a third person
- The murder weapon was a cylindrical object, such as a pipe or flashlight, not a surgical instrument, as asserted by the Coroner
- The physical evidence demonstrated that the crime was a sexual assault
One very large blood spot was present on the wardrobe door (photographs 14, 16). It measured about 1" in diameter at its largest dimension. It was essentially round, showed no beading, and had impinged almost exactly perpendicular to the door, i.e. horizontally and at right angles to the door. This spot could not have come from impact spatter. It is highly improbable that it could have been thrown off a weapon, since so much blood would not have adhered during the back swing for so long a distance and then separated suddenly at just the right moment to deposit as it did. This spot requires an explanation different from the majority of the spots on the doors. It almost certainly came from a bleeding hand, and most probably occurred at a time different from the time that hand was wielding a weapon. The bleeding hand could only have belonged to the attacker. The origin of the injury is dealt with elsewhere, as is supplementary confirmation of the different origin of this spot. It should be noted that this spot is probably not unique in origin, and other spots on the east wall and possibly elsewhere may have had the same origin, but this spot was unique in size and appearance and was consequently selected for more extensive study.
Examination of the mattress top as well as superficial examination of the under sheet and pad on which the victim was murdered shows certain interesting and possibly pertinent facts. Examination of the bedding shows the presence in considerable quantity of a fluid other than blood, most heavily concentrated in the lower portion of the bedding, and forming a large part of the large central bloody area. This fluid was urine, probably voided at or shortly before death. It was probably hypotonic, i.e. less concentrated than the blood, for it appears to have produced hemolysis of the blood corpuscles as it mixed with the blood. On the east side of the bed, visible on the lower sheet, corresponding to the edge of the mattress and just south of the center point of the sheet, is a region that appears bloody, but with very dilute blood. This spot is 9 inches wide at its widest point, the south edge being 3 ft. 3 inches from the south edge of the sheet and the north edge 3 ft. 9 inches from the north edge of the sheet. Examination of this spot visually and with magnification (in the prosecutor's office) showed the blood to be highly dilute and almost certainly hemolyzed. This could happen by mixing the blood with any dilute water solution or water itself, as well as with urine. Its position, shape, and size are most consistent with it having been made by a wet knee placed against the sheet. Inspection of the spot shows that blood was carried laterally with the flow of fluid. Original blood spots are still present, only partially displaced by the diluting fluid. A further observation of the blood pattern on the sheet is significant, which consists of an area approximately 12-18 inches to the right of the wet spot discussed above and 18 inches from the edge of the sheet to the center of the area. This area contained numerous original, undiluted blood spots that had been strongly smeared in the north-south direction, lengthwise on the bed. The area involved was at the exact spot that the attacker must have occupied to intercept the blood spots on the walls as they were intercepted; it lies just where a knee would have been placed to balance him during the wielding of the weapon.
The weapon was short, as shown by the reconstruction diagram. Having fixed the position of the attacker and knowing the position of the victim's head, the length of arc is exactly what would be true of a man's arm wielding a weapon less than one foot in length, i.e. about 36 inches in total. Naturally, the torso and arm lengths influence weapon length calculation, because the distance that can be established is the sum of the arm and weapon length. Even with a short arm, the length of 1 foot covers the available and necessary distance.
Fresh blood on a smooth non-absorptive surface dissolves and washes away rapidly in water. In order to determine how long blood would remain on an expansible metal watch band under these conditions, such a band was daubed liberally with freshly shed blood in two separate experiments. In the first instance the blood was allowed to remain 20 minutes after deposition and the band was then dipped in fresh water and moved slowly back and forth. In the second, the blood was allowed to dry for it hours and treated similarly. In both instances, the blood dissolved rapidly and was essentially gone in less than 1 minute. In the latter experiment, a bit of clot in a recess in the band persisted for about 3 minutes but was washed colorless in less than 2 minutes. In no case could the blood have been grouped after 1 minute in the water, and less than half a minute removed approximately 90% of it. This experiment emphasizes the great importance of the type of surface on which blood is deposited because it is very difficult to remove blood completely from absorptive surfaces. It also is conditioned by the fact that the amount of blood necessary for grouping is far in excess of that necessary for detection, and greater than is required for precipitin tests.
Blood from the same source that was used on the watch band was smeared over the back of the hand. The time was taken at the moment of wetting the band, and the temperature and humidity were recorded: temperature 68°F, humidity 56% relative. Within 2 1/4 minutes, the blood was completely dry.
A series of five cloths, including a variety of cottons, wools, rayons, and silk, were suspended and liquid human blood was thrown against them by means of a brush dipped in blood. It was applied plentifully so that much of it flowed immediately from the garment. The time was taken when the blood was applied and measured until the last drop fell spontaneously from any of the garments.
The condition of the cloths and the three or four drops that fell after the first rapid drainage are illustrated. The three or four drops fell within 2 1/2 minutes, after which no further drainage occurred and the remaining blood dried on the clothing. Shaking of a cloth after application of liberal amounts of blood caused the removal of nearly all of the excess immediately. Thus, whether clothing is shaken by the movement of its wearer or allowed to stand completely quiet, no blood drains after a short interval of time; that which is retained is either absorbed by the cloth (predominate) or dries in a crust on the surface.
Two series of experiments were performed with a variety of objects that would illustrate effects similar to some common weapons:
1. A large bread knife with a roughly triangular blade 8" in length and a breadth at the widest point of 1 1/2"
2. A large monkey wrench 15" in length with a jaw 1 3/4" deep and a maximum opening of 4"
3. A brass bar 11 3/4" in length, 3/4" wide, and 1/8" thick
4. A bar of soft wood 23" long, 1" wide, and 7/16" thick
5. A small ball-peen hammer with a head length of 2 1/2" and a face 3/4" in diameter