‘I would do anything for a handful of dirt,’ he’d moan. ‘Just a spoonful. A crumb.’ He’d sigh dramatically at every sparkling ice core we dissected, but it was all in jest. Anytime something new was brought into the clean lab, his eyes lit up as much as anyone’s.

At the end of Chikondi’s first period of photography, we gathered at his request in the data lab. The three of us sat around the table, tablets and attention at the ready.

Chikondi stood at the sketch board, hands full of styluses, his whole body about to burst with excitement. ‘Draw anything you haven’t seen before up here on the board, and call it out when you do, so we don’t end up “discovering” the same thing twice. Now, at first, everything’s going to be something we haven’t seen before, so we’ll begin with reviewing images as a group until we start getting familiar with the phenotypes that are out there. Okay?’

Jack looked at me. ‘Can you draw for me? I’m shit at drawing.’

‘No,’ I said.

‘Elena, would you—’

Elena’s expression answered his question beyond a shadow of a doubt.

‘So long as it’s got the right number of legs and an approximate body shape, you’ll be fine,’ Chikondi said.

‘But it’s for posterity.’ Jack gestured in protest at the sketch board, which would digitally record, archive, and transmit everything we scribbled on it. ‘I don’t want the history books to say, “This is Jack Vo, trailblazer, Renaissance man, but ultimately, tragically, shit at drawing.”’

Elena gave him a look. ‘You’re being an infant.’

Jack wrinkled his nose at her. ‘You’re being an infant.’

Chikondi handed Jack the first stylus. He passed styluses to me and Elena as well, then set down the remaining fistful on one of the lab benches.

‘Why did you grab so many?’ Elena asked.

‘Well, just in case one breaks,’ Chikondi said. ‘We’re going to be here for a while.’ He tapped his tablet, syncing his screen to the larger display next to the sketch board, then settled into one of the lab chairs. ‘Okay, get ready, everyone. Day one, image one.’

We leaned forward.

Standing on the ice and watching the life forms in motion had been mesmerising. Seeing a snapshot of said same – a dozen or so glowing creatures frozen in time, their forms finally still enough for us to bloodlessly dissect – made us pine.

Jack leapt to his feet, artistic insecurities no match for the siren’s call of raw data. ‘Put a grid on it.’

Chikondi tapped his screen, and a neat net of squares appeared over the image. He got up, too, his rest in the chair short-lived. He looked electric, like if you touched him you’d feel a snap. ‘Okay, okay, ah – first row, first column—’

Jack nodded and began to draw. ‘AnA, yeah?’ By this, he meant Annelid Approximate, one of OCA’s many official classifications. You don’t want to call an alien creature a fish or a spider in a field research context. It may look like an animal back home, and may even behave like an animal back home, but it’s not the same thing, and shoving everything we find out here into the categories we have on Earth is a dangerous trap. You have to give some kind of name to the things you find, though, and as taxonomy is the sort of long-game activity best done back home, we survey teams use simple acronyms based on terrestrial phyla, to help us visually sort things until proper classification can be determined. So, because it’s not very fancy to say ‘worm-like,’ we say Annelid (e.g. earthworm) Approximate: AnA. You can find this acronym on the same list as Avian Reptile Approximate (AvA), Amphibian Approximate (AA), Mammal Approximate (MA), and so on – plus the ever-exciting ‘NP’ for ‘New Phyla’. Everybody wants to find an NP.

Chikondi got blindingly close to the monitor as he weighed Jack’s assessment of the creature in the top left grid. He thought for a minute, then shook his head. ‘It’s not segmented, it’s smooth. And stocky. I say CA.’ Cnidarian Approximate, the phyla that includes sea cucumbers.

‘Hold on.’ Elena joined the fray. ‘It’s got feet.’

‘Where?’ Chikondi asked.

Elena pointed at a different grid. ‘This is the same species, right?’ She made a pulling motion on the monitor, zooming in. ‘Look. Isn’t that a foot?’

I got up to look, too. We all squinted at the tiny blobs sticking out of the larger blob.

‘Hard to say,’ Jack said.

‘It’s in motion,’ I said. ‘We need a clearer image.’

‘I swear that’s a foot,’ Elena said. ‘Or a digit of some kind.’

‘Cnidarians have feet, so CA would still be accurate,’ Chikondi said. ‘Although . . .’ He shook his head with the kind of frustrated puzzlement every scientist longs for. He zoomed in closer and frowned. ‘Does it have bones?’

We leaned even further forward with cartoonish synchrony. We looked, and we looked, and we looked, the pixels somehow becoming less clear with each second that went by.

‘Mark it as inconclusive until we see more specimens,’ Chikondi said to Jack.

Jack began to write on the sketch board below his drawing of the creature.

Elena looked at Jack’s handiwork. ‘You really are shit at drawing.’

He casually gave her the finger with his free hand as he wrote: CA0001 (incon.).

So it went for two hours, until all fuel for categorical bickering had been spent. A menagerie of crudely sketched body types filled the board – thirteen suspected new species in total. Thirteen unique animals with their own lives and stories to tell. Thirteen things no human before us had ever seen.

The camera trap is one of the most humane inventions in the ecological survey toolbox, and it excels when paired with its best friend, image recognition software. In the old days, scientists had to manually review every image they took of a jungle road or a wildlife corridor, painstakingly tagging each file with whether it contained elephants or bears or whatever biological quarry they were after. It took a tremendous number of work hours, and often had to be completed by volunteers, simply because there was never enough funding to actually pay the number of people such a task would require. Software that could do the work in a fraction of the time was a godsend, and its development revolutionised the field. There’s only one problem with that approach: you have to teach the software what to look for. If you’re dealing with creatures nobody’s ever seen, you can’t do that, not at first. The only way forward is old school.

‘Okay,’ Chikondi said energetically, tapping his tablet. ‘On to day one, image two.’

‘How much time lapsed between each photo capture?’ Jack asked.

‘Two minutes,’ Chikondi said.

Jack exhaled and cracked his neck. I clapped him on the shoulder in solidarity. Elena smiled silently and folded her arms, looking ready – no, eager to forgo sleep for this. She was made for this work. We all were.

A scholar could review the reports we made on Aecor in days, maybe weeks. For the layman, our discoveries can be summed up as follows:

We found an animal with a previously unknown method of propulsion: Tubuspiscis quesadae, which looks like a sport sock with the entire toe cut out, if that sock were made of the flesh of a jellyfish instead of fabric. It squeezes the sides of its hollow body to propel itself through the water, gathering nutrients in the dense fur of filters that coats its inner side.