Each email server (using the SMTP protocol) adds a
Received: header specifying who the server is, from where it got the email, and when. In your case, one of the servers internal to the Gmail architecture received the email, apparently through some HTTP interface, and that machine claims to have the specified IP address.
Let's see an example. "Alice Example" (Gmail user, address
firstname.lastname@example.org) sends an email to here brother "Bob Example", owner of the
example.com domain and the
email@example.com email address. Bob finds the following header in his mailbox:
X-Spam-Checker-Version: SpamAssassin 3.3.2 (2011-06-06) on smtp.example.com
X-Spam-Status: No, score=-0.7 required=5.0 tests=FREEMAIL_FROM,HTML_MESSAGE,
Received: from mail-ee0-f54.google.com (mail-ee0-f54.google.com [220.127.116.11])
by smtp.example.com (Postfix) with ESMTP id 56C8520390
for <firstname.lastname@example.org>; Thu, 5 Sep 2013 14:21:21 +0200 (CEST)
Received: by mail-ee0-f54.google.com with SMTP id e53so864366eek.13
for <email@example.com>; Thu, 05 Sep 2013 05:21:20 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
X-Received: by 10.14.107.68 with SMTP id n44mr13246148eeg.26.1378383680460;
Thu, 05 Sep 2013 05:21:20 -0700 (PDT)
Received: by 10.15.90.131 with HTTP; Thu, 5 Sep 2013 05:21:20 -0700 (PDT)
Date: Thu, 5 Sep 2013 08:21:20 -0400
From: Alice Example <firstname.lastname@example.org>
To: Bob Example <email@example.com>
Content-Type: multipart/alternative; boundary=001a11c29ad2a386e504e5a1f573
The headers are added on top by each successive server, so the first
Received: (and headers which come before that) was added by the last SMTP server, i.e.
smtp.example.com. The contents of that header mean that this server received the email from another server who claimed to be named
mail-ee0-f54.google.com, and had IP address 18.104.22.168, and that IP address resolves (through reverse DNS) to
mail-ee0.f54.google.com (the same name as the claimed name, which is good).
Received: header was the one which was first added; it tells us that (apparently) the first SMTP-conscious server who received the email got it over HTTP (that is, a custom protocol which uses HTTP as transport) and that server has the name "10.15.90.131", which is an IP address. We may thus surmise that the Web server to which Alice's browser connected has, within Gmail's network, a private IP address (10.15.90.131). This does not tell us much, because it is a private IPv4 address so it has no meaning outside of Gmail's network.
In any case, that IP address does not indicate much about the IP address of Alice's machine (where she runs her browser). Possibly, Gmail's systems are spread throughout the World and Alice's connection was routed to a Web server in the "same area" (for improved performance). By analysing many headers from many emails, we might rebuild a sort of mapping from geographical area to internal IP address, giving a clue about current Alice's whereabouts. However, this would rely on undocumented assumptions about Gmail's network, and would be crude. Note that this first internal SMTP server uses the PDT time zone, corresponding to North America's West coast, while (in that case) Alice's machine is in North America's East cost, more than 4000 km away from the Pacific.
Summary: the IP addresses you see relate to the servers in Gmail's systems, and are not really related to the client's address. It is the internal address of the server which first received the email. What is relevant for privacy is that the same server did not specify the IP address of the client; the address you see is not a "scrambled" client's IP address; it is another unrelated address.