diff --git a/app.py b/app.py
index 3108b9d..7827653 100644
--- a/app.py
+++ b/app.py
@@ -36,6 +36,12 @@ def report():
         n_ts += 1
     return jsonify({ "success": True, "timeseries": n_ts, "datapoints": n_dp })
 
+@app.route("/ts/<id>")
+def get_timeseries(id):
+    breakpoint()
+    ts = LTS(id=id)
+    return jsonify({"description": ts.description, "data": ts.data})
+
 def verify_node(d):
     node = d["auth"]["node"]
     timestamp = d["auth"]["timestamp"]
diff --git a/ltsdb_json.py b/ltsdb_json.py
index 6eb2f3a..bb155cd 100644
--- a/ltsdb_json.py
+++ b/ltsdb_json.py
@@ -11,19 +11,21 @@ class LTS:
     base_dir = "data"
     limit = 1000
 
-    def __init__(self, description):
-        # Oh, I think we need to be able to load by hash, too
-        canonical_description = {x: description[x] for x in sorted(description.keys())}
-        self.description = canonical_description
-        serialized_description = json.dumps(canonical_description)
-        m = hashlib.sha256()
-        m.update(bytes(serialized_description, encoding="UTF-8"))
-        self.filename = self.base_dir + "/" + m.hexdigest()
+    def __init__(self, description=None, id=None):
+        if description:
+            canonical_description = {x: description[x] for x in sorted(description.keys())}
+            self.description = canonical_description
+            serialized_description = json.dumps(canonical_description)
+            m = hashlib.sha256()
+            m.update(bytes(serialized_description, encoding="UTF-8"))
+            id = m.hexdigest()
+        self.filename = self.base_dir + "/" + id
         try:
             with open(self.filename, "r") as fh:
                 fcntl.flock(fh, fcntl.LOCK_SH)
                 d = json.load(fh)
             self.new = False
+            self.description = d["description"]
             self.data = d["data"]
         except FileNotFoundError as e:
             self.new = True