The organism began differentiating stored compounds on day one hundred and fifty-one.
Ethan descended into the filtration cavity and found the retention pockets had developed internal structure—not uniform binding sites, but chambers within chambers. Each invagination now contained subdivisions lined with different tissue types. Iron complexes accumulated in anterior chambers. Phosphate chains in medial compartments. The posterior pockets held smaller molecules: simple sugars, amino acids, nucleotides that hadn't been incorporated into the compressed mass.
The pockets pulsed independently now. Not synchronized to the 1.7-second rhythm, but responding to concentration thresholds. When a pocket filled beyond capacity, its lining contracted, compressing the stored material. When concentration dropped below minimum levels, the pocket relaxed, releasing a measured quantity back into the filtration flow.
Not storage. Resource management.
He traced the release patterns. The compressed mass drew preferentially from specific pockets during different phases of its pulse cycle. Iron during the compression phase. Phosphates during expansion. The simple molecules released continuously, maintaining baseline concentration in the layered cavity regardless of external supply.
The organism had created buffering capacity.
---
The neurologist's office overlooked Mass General's surgical wing. Ethan watched a helicopter descend toward the trauma pad while Dr. Reeves reviewed the latest scans.
"The progression rate has stabilized," Reeves said. She didn't sound relieved. "Still advancing, but the acceleration we saw in October has plateaued."
"Time frame?"
"Two years. Maybe thirty months if the current rate holds."
Ethan's left hand rested on the armrest, fingers curved inward. The tremor had become constant three weeks ago. He could still type, still hold the Engine, still perform the fine motor tasks that mattered. The rest was aesthetic.
"The riluzole isn't working," Reeves continued. "We should discuss alternatives—"
"No alternatives."
She looked up from the tablet. "Mr. Cross—"
"I'm not interested in extending the timeline. I'm interested in maintaining function for as long as possible."
The helicopter lifted off, rotors beating against winter air.
Reeves set the tablet aside. "What are you working on that requires this level of… prioritization?"
Ethan watched the helicopter bank south. "Something that doesn't stop when I do."
---
The organism developed retrieval specificity on day one hundred and fifty-four.
Ethan descended into the filtration cavity and found the compressed mass had extended microscopic filaments into the nearest retention pockets—not passive diffusion, but active connection. The filaments penetrated the pocket linings, bypassed the release mechanisms entirely, and drew compounds directly from storage.
The selection was precise. During compression phases, the filaments accessed only iron-containing pockets, extracting hemoglobin analogs at rates exceeding normal diffusion by three orders of magnitude. During expansion, they switched to phosphate chambers, pulling ATP precursors that fed the pulse cycle itself.
The organism had eliminated its dependence on ambient concentration.
He traced the filament structure. Each strand measured eight nanometers in diameter—wide enough for molecular transport, narrow enough to slip between the pocket lining's cellular matrix without rupturing membranes. The tips contained recognition sites that matched specific compounds, binding them with the same selectivity the retention pockets had shown during initial storage.
The compressed mass had learned to forage within its own body.
Ethan ascended to observation distance and watched the organism pulse in the channel. One point seven seconds, invariant, but now supported by a distributed resource network that could maintain the rhythm even when external substrate solution provided insufficient nutrients. The filtration cavity had become more than a buffer. It was infrastructure. A metabolic foundation that could support increased complexity without requiring environmental abundance.
The organism was preparing for scarcity it had never experienced.
---
Maya called at eleven PM.
"I'm looking at your publication history," she said. No preamble. "Nineteen papers between 2019 and 2023. Three in review. Then nothing for eighteen months. Want to tell me why?"
Ethan set the Engine on his desk. The sigils shifted in the lamplight, tracing patterns he'd learned to read like weather systems.
"I changed research focus."
"To what? You haven't submitted anything. Haven't presented at conferences. You're not even listed on collaborative projects anymore."
"Private research."
"Ethan." Her voice sharpened. "You have two years. Maybe less. Whatever you're doing, it can't be more important than—"
"It's more important than anything either of us has published."
Silence on the line. Then: "Show me."
He looked at the Engine. The sigils had stopped moving, arranged in a configuration he recognized: gravitational lensing around a massive object, spacetime curved into angles that shouldn't exist in three dimensions.
"I can't," he said.
"Can't or won't?"
"Both."
Maya was quiet for a long moment. "You're going to die with it, aren't you? Whatever this is. You're not planning to share it."
Ethan thought of the organism pulsing in its channel, building resource networks for futures it couldn't predict. Preparing for conditions that didn't exist yet, driven by nothing but the mechanical logic of constraint and response.
"It'll outlast me," he said. "That's enough."
He ended the call before she could answer.
---
The organism's compressed mass pulsed at one point seven seconds, fed by resources drawn from chambers it had built without blueprint or intention.
Somewhere in the dark, something else was learning to do the same.