Seemed like quite a lot to me.
I looked next at 2005. Only six women in the corresponding three age brackets had died. In 2006, there had been only four such deaths.
2006 was the latest year for which stats had been collected so I went back in time. In 2003, two women in the age range had died. 2002 had been a particularly good year for young women with no deaths recorded. 2001, on the other hand, recorded eleven deaths.
I went further back. 2000 had seen six mortalities, there were just two in 1999, but an impressive ten in 1998.1997 saw a modest two, as did 1996, but – would you believe it – in 1995 eight women had met a premature end.
I went back as far as the table did, to 1983. I'm no statistician but even I could see a pattern emerging. Every three years a modest but significant blip appeared in the female death rates. Now, what the hell did that mean and, more importantly, why had no one spotted it before?
I looked again at the total column, to see if the pattern was reflected there. The number of total female deaths on Shetland ranged quite considerably from just 86 in 2003 to 154 in 1997. I looked for quite some time but there was no discernible three-year pattern that I could see; the spread of deaths, the difference in the numbers, appeared totally random. Whatever was going on in the three age groups representing young adult women was being hidden by the larger numbers involved in the female population as a whole. When you added male deaths into the equation, the chances of someone spotting what I'd just seen were non-existent.
Which might explain why no keen-eyed statistician at the General Register Office had spotted the anomaly. Taking the Shetland population as a whole, nothing was happening; and, when you considered that the death rate on Shetland is actually lower than the rest of Scotland, there'd be no reason for anyone to look more closely at the figures. The numbers were simply too small to show up on anything other than a very focused search.
I sat back to think about it.
I'd gone looking for one batch; I'd found seven. At least seven years in which the female death rate leaped way beyond the norm. If these figures were shown to someone in authority, surely they'd be enough to convince them that something odd was going on. Unfortunately, I had no idea whom to approach. Whilst I couldn't believe the entire Northern Constabulary was corrupt, without Dana how would I know whom to trust and whom not? More significantly, if some of these deaths were suspicious (or – not to put too fine a point on it – if they hadn't actually occurred), then how could some very senior people at this hospital not be involved? Could I rely on anyone to back me up? I decided I needed more details. Who were these dead women? How had they died? I'd start with 2004, the year Melissa had supposedly died.
I left the Internet, went into the hospital's own records and called up details of mortalities for 2004. There were 106 female deaths in total that year, out of which I was looking for just twelve. This was going to take some time and I was still woozy from the sedative Gifford had given me.
Fortunately, the list of mortalities included a name and a date of birth. It took about thirty minutes – of jumping out of my skin every time I heard a noise in the corridor – but eventually I had my list of twelve women, aged between 20 and 34, who died that year.
I scribbled their names, ages and an abbreviated cause of death on a notepad I found on the desk.
Melissa Gair -32 -breast cancer
Kirsten Hawick -29 -riding accident
Heather Paterson -28 -suicide
Kate Innes -23 -breast cancer
Jacqueline Ross -33 -eclampsia
Rachel Gibb -21 -car accident
Joanna Buchan -24 -drowning
Vivian Elrick -27 -suicide
Olivia Birnie -33 -heart disease
Laura Pendry -27 -cervical cancer
Caitlin Corrigan -22 -drowning
Phoebe Jones -20 -suicide
I stared at the list for five, ten minutes, looking for anything out of the ordinary. There was nothing, really, except that there seemed too many of them. Otherwise, the causes of death were exactly what you might expect. When young women die, it's usually as a result of an accident or deliberate self-harm. Other than that, you might expect a few cases of heart disease and cancer and, occasionally, problems associated with childbirth.
I looked back to my first list, the one I'd printed off the General Register site. A very crude calculation told me that when you removed the blip years from the equation, the average number of young women to die on Shetland each year was 3.1. Using just the blip years, the average leaped to ten. Every three years, six or seven more women than usual were dying.
Was it remotely possible to fake that number of deaths; to spirit these women away, keep them alive for a further year, before murdering them as brutally as Melissa had been murdered? And – big question coming up – had they, like Melissa, given birth shortly before death?
I turned back to the list of twelve women who'd died in 2004. Melissa and Kirsten hadn't died natural deaths, I was sure of that now. But which of the others had shared their fate? Vivian? Phoebe? Kate? Which of these women had been abducted, kept prisoner for the better part of a year, given birth alone and terrified? What had been their biggest fear at the end – fear for themselves or for what was going to happen to their babies?
A harvest of babies. At last I'd said it. It must have been bouncing around in the back of my mind ever since the post-mortem, when we'd learned that my lady from the peat had had a baby. What had happened to that baby, I'd asked myself immediately. In Richard's study, discovering that one of the runes meant Harvest had almost got me there but it had taken Jenny's passing comment about batches to give me the shove I needed.
OK, think, Tora, think. If these women were being abducted, they had to be kept somewhere; somewhere secure and out-of-the- way, but local all the same. They were buried here – in my backyard, for God's sake – so they couldn't have left the islands. It needed to be somewhere with medical facilities, where babies could be safely delivered. Jesus! It was obvious.
I started typing again and brought up the obs and gynae pages. I'd produced this list once before, the day after I'd found Melissa: the details of all births on the islands between March and August 2005 – the time Melissa's baby would have been born. I printed it off and sat looking at it, refreshing my memory. One hundred and forty births. According to Dana, most of the women named had been accounted for, found alive and well, but I already knew I was dealing with clever, impressively resourced people. If you could fake a death in a modern hospital, you could falsify pretty much anything.
I went down the list, marking entries as I went. Soon, every birth that had taken place on Tronal was highlighted in yellow. I was looking for six or seven; I found four. It was too few to suggest an easy answer; and yet Tronal was the ideal place: remote enough to offer privacy but accessible for those with their own boats and able to cope with the difficult navigation. It had a modern maternity facility and its own resident obstetrician. I realized, with a sinking heart, that it also had a qualified anaesthetist well within commuting distance.
Oh, Christ!
My father-in-law was involved in the facility on Tronal. He had to be; that was where he went when he left the house most days. I remembered what Stephen Renney had said about Melissa being heavily anaesthetized before she was killed and nausea reared. Richard had been Medical Director at the Franklin Stone before passing on the reins to his protege, Kenn Gifford. If deaths were being faked at this hospital then the medical directors were ideally placed to oversee it happening.
Suddenly I was sure Richard was involved. Probably Kenn, too. And Dana and I had both had our doubts about Andy Dunn. One of them had watched Duncan and me set off in our dinghy, believing that I wouldn't survive the trip. They'd conspired to murder me. And they would try again.