diff --git a/parser.py b/parser.py
index cc34439..3d76a1a 100755
--- a/parser.py
+++ b/parser.py
@@ -24,6 +24,7 @@
 import os
 import re
 import shutil
+from collections import namedtuple
 
 def read_json_from_js_file(filename):
     """Reads the contents of a Twitter-produced .js file into a dictionary."""
@@ -42,35 +43,38 @@ def extract_username(account_js_filename):
     account = read_json_from_js_file(account_js_filename)
     return account[0]['account']['username']
 
+ExtractedTweet = namedtuple('ExtractedTweet', 'id reply_to_self reply_to_id body')
+
 def tweet_json_to_markdown(tweet, username, archive_media_folder, output_media_folder_name, tweet_icon_path):
     """Converts a JSON-format tweet into markdown. Returns tuple of timestamp and markdown."""
+
     tweet = tweet['tweet']
     timestamp_str = tweet['created_at']
     timestamp = int(round(datetime.datetime.strptime(timestamp_str, '%a %b %d %X %z %Y').timestamp())) # Example: Tue Mar 19 14:05:17 +0000 2019
     body = tweet['full_text']
     tweet_id_str = tweet['id_str']
+
     # replace t.co URLs with their original versions
     if 'entities' in tweet and 'urls' in tweet['entities']:
         for url in tweet['entities']['urls']:
             if 'url' in url and 'expanded_url' in url:
                 body = body.replace(url['url'], url['expanded_url'])
+
     # if the tweet is a reply, construct a header that links the names of the accounts being replied to the tweet being replied to
-    header = ''
+    in_reply_to_status_id = ''
+    reply_to_self = False
     if 'in_reply_to_status_id' in tweet:
+        in_reply_to_status_id = tweet['in_reply_to_status_id']
         # match and remove all occurences of '@username ' at the start of the body
         replying_to = re.match(r'^(@[0-9A-Za-z_]* )*', body)[0]
         if replying_to:
-            body = body[len(replying_to):]
+            # some old tweets lack 'in_reply_to_screen_name': use it if present, otherwise fall back to names[0]
+            in_reply_to_screen_name = tweet['in_reply_to_screen_name'] if 'in_reply_to_screen_name' in tweet else replying_to.split()[0]
+            body = f'[{replying_to[:-1]}](https://twitter.com/{in_reply_to_screen_name}/status/{in_reply_to_status_id}) ' + body[len(replying_to):]
         else:
-            # no '@username ' in the body: we're replying to self
-            replying_to = f'@{username}'
-        names = replying_to.split()
-        # some old tweets lack 'in_reply_to_screen_name': use it if present, otherwise fall back to names[0]
-        in_reply_to_screen_name = tweet['in_reply_to_screen_name'] if 'in_reply_to_screen_name' in tweet else names[0]
-        # create a list of names of the form '@name1, @name2 and @name3' - or just '@name1' if there is only one name
-        name_list = ', '.join(names[:-1]) + (f' and {names[-1]}' if len(names) > 1 else names[0])
-        in_reply_to_status_id = tweet['in_reply_to_status_id']
-        header += f'Replying to [{name_list}](https://twitter.com/{in_reply_to_screen_name}/status/{in_reply_to_status_id})\n\n'
+            # decide later if a header should be written
+            reply_to_self = True
+
     # replace image URLs with markdown image links to local files
     if 'entities' in tweet and 'media' in tweet['entities'] and 'extended_entities' in tweet and 'media' in tweet['extended_entities']:
         original_url = tweet['entities']['media'][0]['url']
@@ -100,11 +104,18 @@ def tweet_json_to_markdown(tweet, username, archive_media_folder, output_media_f
                         print(f'Warning: missing local file: {local_filename}. Using original link instead: {original_url} (expands to {original_expanded_url})')
                         markdown += f'![]({original_url})'
         body = body.replace(original_url, markdown)
+
     # make the body a quote
     body = '> ' + '\n> '.join(body.splitlines())
+
+    # give every tweet an anchor
+    body = f'<a name="s{tweet_id_str}"></a>\n' + body
+
     # append the original Twitter URL as a link
-    body = header + body + f'\n\n<img src="{tweet_icon_path}" width="12" /> [{timestamp_str}](https://twitter.com/{username}/status/{tweet_id_str})'
-    return timestamp, body
+    body += f'\n\n<img src="{tweet_icon_path}" width="12" /> [{timestamp_str}](https://twitter.com/{username}/status/{tweet_id_str})'
+
+    return timestamp, ExtractedTweet(id=tweet_id_str, reply_to_self=reply_to_self, reply_to_id=in_reply_to_status_id, body=body)
+
 
 def main():
 
@@ -141,6 +152,8 @@ def main():
     if not os.path.isfile(tweet_icon_path):
         shutil.copy('assets/images/favicon.ico', tweet_icon_path);
 
+    connected_tweets = {}
+
     # Parse the tweets
     username = extract_username(account_js_filename)
     tweets_markdown = []
@@ -148,7 +161,10 @@ def main():
         print(f'Parsing {tweets_js_filename}...')
         json = read_json_from_js_file(tweets_js_filename)
         tweets_markdown += [tweet_json_to_markdown(tweet, username, archive_media_folder, output_media_folder_name, tweet_icon_path) for tweet in json]
-    print(f'Parsed {len(tweets_markdown)} tweets and replies by {username}.')
+
+        for tweet in json:
+            if 'in_reply_to_status_id' in tweet['tweet']:
+                connected_tweets[tweet['tweet']['id']] = tweet['tweet']
 
     # Sort tweets with oldest first
     tweets_markdown.sort(key=lambda tup: tup[0])
@@ -162,9 +178,37 @@ def main():
         filename = filename = f'{dt.year}-{dt.month:02}-01-Tweet-Archive-{dt.year}-{dt.month:02}.md' # change to group by day or year or timestamp
         grouped_tweets[filename].append(md)
 
+    tweet_files = {}
+    for filename, md in grouped_tweets.items():
+        for t in md:
+            tweet_files[t.id] = filename
+
+    forward_refs = defaultdict(list)
+    for filename, md in grouped_tweets.items():
+        for t in md:
+            if t.reply_to_id and t.reply_to_id in tweet_files:
+                forward_refs[t.reply_to_id].append(t.id)
+
     # Write into files
     for filename, md in grouped_tweets.items():
-        md_string =  '\n\n----\n\n'.join(md)
+        md_string = ''
+        prev_id = ''
+        for t in md:
+            # got successors?
+            first_forward = True
+            for s in forward_refs[prev_id]:
+                if s != t.id:
+                    md_string += '\n\n' if first_forward else ''
+                    first_forward = False
+                    md_string += f' [↓]({tweet_files[s]}#s{s})'
+            if (not t.reply_to_id or t.reply_to_id != prev_id):
+                # not a reply to the last output tweet: write a horizontal rule
+                md_string += '\n\n----\n' if md_string else ''
+                if(t.reply_to_self):
+                    # a reply-to-self, but not a reply to the most recently output tweet: add a header
+                    md_string += f'\n[↑]({tweet_files[t.reply_to_id]}#s{t.reply_to_id})' if t.reply_to_id in tweet_files else '↑'
+            md_string += f'\n{t.body}'
+            prev_id = t.id
         with open(filename, 'w', encoding='utf-8') as f:
             f.write(md_string)
     print(f'Wrote tweets to *.md, with images and video embedded from {output_media_folder_name}')