A murder conviction normally begins with a body. Twelve entries in this archive describe killings where one was never recovered, and reading them together shows how much of a victim count is really a question about evidence rather than about what happened.
Robert Zarinsky's case is the one that established the point. He was convicted in 1975 of the murder of Rosemary Calandriello, whose body was never found, in one of the first New Jersey convictions obtained without one, and sentenced to life imprisonment. He was acquitted in later trials over two other deaths, and died in prison in 2008. That is a court deciding that the absence of a body is not the same as the absence of a death.
Most of the entries here work the other way, and the pattern is consistent. Where a body was recovered, there is a conviction. Where one was not, there is a number in a range.
Ma Yong's case shows it inside a single trial. Twelve killings were attributed to him and the court confined its findings to those twelve, plus participation in five for his co-defendant Duan. Two bodies were never found. Vasile Tcaciuc has twenty-one proved and twenty-six alleged; the remaining bodies were never recovered, and twenty-six is the figure most often given. He confessed in custody in 1935 and no trial or execution record is accessible, so nothing against him was ever tested. Darren Dee Oneall has one conviction against six or more attributed, and in several of them the remains were never recovered, no charge was brought in any of those.
Stanisław Modzelewski's entry states the reasoning outright. A killing was suspected and never proved, because the body was never found, and for that reason it is not in the roster's count. That is the archive applying its own rule in a single sentence: the count follows what a court could establish, and a missing body usually means it could not.
Some of these are not ambiguous at all about what happened, only about what could be charged. John Coombes killed Michael Speirani, twenty, by running him over with a motorboat in February 1984, and the body was never recovered. Peter Kudzinowski's third victim was never found. Henri Désiré Landru's eleven victims disappeared into a stove at Gambais, and no body was ever recovered from any of them, he was convicted of all eleven and guillotined, on physical traces and paperwork rather than remains.
Devernon LeGrand's victims' bodies were never recovered either, and his roster figure is twelve or more. He was convicted with his son in 1977 of two of them.
Two cases in this group show what the gap costs when the killer is never tried for murder at all.
David Parker Ray was convicted of kidnapping and sexual torture, never of murder, and no body was ever found. His victim count in this archive is recorded as none established, which is the honest entry for a man widely believed to have killed and never convicted of it. Nicholas Wiley has three, one of whose bodies was never found, and one of the women he killed was five months pregnant.
Joel Rifkin's is the saddest version. He killed nine women and three others are recorded only as a first name or a number, of those, two were never found. The anonymity and the missing remains are the same problem seen twice: nobody reported them, and nothing was recovered to identify.
What the twelve say together is that a victim count is not a body count. It is a count of deaths a court could prove, and the single most common reason it could not is that there was nothing to examine.
Which also means the ranges in this archive, twenty-one to twenty-six, one to six, nine to seventeen, are not sloppiness. In most cases the gap between the two numbers is a list of people whose bodies are still somewhere.